Prediction method of earthquake peak acceleration based on second-order neuron deep neural network

A technology of deep neural network and peak acceleration, applied in neural learning methods, biological neural network models, seismology, etc., can solve the problems of low prediction accuracy of earthquake motion, and achieve the effects of good applicability, improved accuracy, and high accuracy

Active Publication Date: 2022-03-11
HARBIN INST OF TECH
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the technical problem of low precision of earthquake prediction, and provide a method for prediction of peak acceleration of earthquake based on second-order neuron deep neural network

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Prediction method of earthquake peak acceleration based on second-order neuron deep neural network
  • Prediction method of earthquake peak acceleration based on second-order neuron deep neural network
  • Prediction method of earthquake peak acceleration based on second-order neuron deep neural network

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0038] Specific implementation mode one: the method for predicting the peak acceleration of earthquake motion based on the second-order neuron deep neural network in this implementation mode is implemented according to the following steps:

[0039] Step 1: Collect earthquake records and create a data set:

[0040] Select the magnitude (M), projection distance (R JB ), shear wave velocity (V S30 ), region (Region), cover layer thickness (Z 1 ), fault type (Fault Type) and period (T) are the input parameters, and the corresponding peak acceleration of the ground motion is the target parameter, and the values ​​of the input parameter and the target parameter are between -0.5 and 0.5 through the standardization method, and the ground motion data set;

[0041] Step 2: Build a deep neural network with second-order neurons:

[0042] A deep neural network consisting of three hidden layers is established. The neurons are all second-order elements. The hyperbolic tangent function (T...

specific Embodiment approach 2

[0059] Embodiment 2: This embodiment is different from Embodiment 1 in that the earthquake records in step 1 are selected from the NGA-West2 database.

specific Embodiment approach 3

[0060] Specific embodiment three: the difference between this embodiment and specific embodiment one or two is that the magnitude (M) in step 1 takes the natural logarithmic value, and the projection distance (R JB ) takes the natural logarithm value.

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The method for predicting the peak acceleration of earthquake motion based on the deep neural network of second-order neurons belongs to the field of earthquake engineering, and it aims to solve the technical problem of low accuracy of earthquake motion prediction. Prediction method of peak ground acceleration: 1. Select magnitude, projection distance, shear wave velocity, area, overburden layer thickness, fault type and cycle as input parameters in the data set, and the corresponding peak ground acceleration as target parameter; 2. A deep neural network consisting of three hidden layers is established. The neurons are all second-order elements. The hyperbolic tangent function is used as the activation function. The mean square error function and Adam adaptive optimization function are used for backpropagation. The mean absolute error function is Evaluation function; 3. Deep neural network model training; 4. Peak acceleration prediction. The invention adopts a multi-input structure and a second-order neuron network, which can not only improve the accuracy of predicting the peak acceleration of earthquake, but also ensure the applicability of the deep neural network model.

Description

technical field [0001] The invention belongs to the field of earthquake engineering, and in particular relates to a method for predicting the peak acceleration of earthquake motion based on a deep neural network of second-order neurons. Background technique [0002] From ancient times to the present, every major earthquake has caused immeasurable losses to the society. In order to reduce and control the losses caused by earthquakes, it is the most important measure to carry out anti-seismic fortification of buildings, and the basis for anti-seismic fortification is earthquake Hazard Analysis. In the seismic hazard analysis, in order to evaluate the seismic intensity through different parameters, it is very important to establish the earthquake motion prediction equation. [0003] At present, the traditional prediction method is mainly to perform empirical regression based on the existing seismic records. This method has good prediction accuracy and uniform distribution of r...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G01V1/30G06K9/62
CPCG01V1/307G06N3/084G06N3/045G06F18/241
Inventor 籍多发翟长海李晨曦温卫平
Owner HARBIN INST OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products