Unlock instant, AI-driven research and patent intelligence for your innovation.

A Fan Noise Prediction Method Based on Neural Network Optimization

A technology of neural network and prediction method, which is applied in the field of noise prediction, can solve problems affecting prediction accuracy, long training time, and low training accuracy, and achieve the goals of improving training accuracy, improving generalization, and avoiding over-fitting and under-fitting Effect

Active Publication Date: 2022-03-01
ZHUZHOU LINCE GRP
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

For the number of input neurons that has been determined, if the number of neurons in the hidden layer is small, the training accuracy will be low, and underfitting will occur. If the number of neurons in the hidden layer is large, the training accuracy will be high, the training time will be long, and overfitting will occur, which will affect the prediction accuracy.

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
  • A Fan Noise Prediction Method Based on Neural Network Optimization
  • A Fan Noise Prediction Method Based on Neural Network Optimization
  • A Fan Noise Prediction Method Based on Neural Network Optimization

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0065] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings, but the present invention can be implemented in various ways defined and covered by the claims.

[0066] Such as figure 1 A fan noise prediction method with optimized neural network is shown, including steps:

[0067] S1. Collect fan performance parameters and geometric parameters;

[0068] There are 6 performance parameters, including: 1-flow, 2-total pressure, 3-speed, 4-power, 5-efficiency, 6-exit wind speed; there are 10 geometric parameters, including: 7-impeller diameter, 8-hub Ratio, 9-blade root installation angle, 10-blade root chord length, 11-blade middle installation angle, 12-blade middle chord length, 13-blade tip installation angle, 14-blade tip chord length, 15-blade number, 16 - Number of guide vanes.

[0069] S2. Analyze the status quo of fan performance parameters;

[0070] Perform performance analysis on the fan parameters teste...

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 invention provides a fan noise prediction method with an optimized neural network. The invention mainly improves the prediction accuracy of the fan noise and the generalization ability of the neural network through joint control of the number of input neurons and the number of hidden layer neurons. The invention sorts the importance of the influence of the input parameters in the fan samples on the output parameters based on the correlation analysis, and determines the range of neurons in the input layer and the optimal number of neurons in the input layer according to the training accuracy and prediction accuracy. The correlation analysis is used to effectively reduce the number of input neurons and reduce the difficulty of constructing the optimal neural network structure. The invention utilizes the number of neurons in the optimal hidden layer to determine the optimal neural network structure, effectively avoids over-fitting and under-fitting, improves the training accuracy and also improves the prediction accuracy and generalization ability.

Description

technical field [0001] The invention relates to the field of noise prediction, in particular to a fan noise prediction method with optimized neural network. Background technique [0002] There are many types of fans and a wide range of applications. Different types of fans, fans in different applications, and fans with different functions have different requirements for allowable noise, which can range from more than ten decibels to nearly one hundred decibels. There are many sources of fan noise, such as: blade rotation, blade vortex, turbulent flow, and resonance with the duct shell. These are all fixed noises of the fan. In addition, there are abnormal noises caused by external assembly and maintenance. Usually there are certain standards for the tolerance of fan noise in specific occasions, so it is particularly important to accurately predict fan noise. [0003] Neural network is a mathematical model that uses a structure similar to that of brain synaptic connections ...

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): G06Q10/04G06N3/02
CPCG06N3/02G06Q10/04
Inventor 刘梦安杨奇阳吉初翟方志侯志泉屈小章
Owner ZHUZHOU LINCE GRP