Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Hardware Realization Method and System of Artificial Neural Network Algorithm

An artificial neural network and neural network technology, applied in the hardware implementation method of artificial neural network algorithm and its system field, can solve the problems of expensive hardware resource occupation and excessive size, and achieve the effect of low cost and saving hardware resources

Inactive Publication Date: 2011-12-07
UNIV OF SHANGHAI FOR SCI & TECH
View PDF0 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] The object of the present invention is to provide a kind of hardware realization method and system thereof of artificial neural network algorithm, overcome the expensive of existing hardware realization technology and take too much defect of hardware resource, realize with VHDL, be convenient to hardware realization and can save hardware resource

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
  • Hardware Realization Method and System of Artificial Neural Network Algorithm
  • Hardware Realization Method and System of Artificial Neural Network Algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0023] In order to make the technical means, creative features, goals and effects achieved by the present invention easy to understand, the present invention will be further described below in conjunction with specific illustrations.

[0024] Such as figure 1 , figure 2 As shown, an artificial neural network algorithm system includes a top-level module xor_s, and the top-level module xor_s is composed of several neuron modules NY3_s. The neuron module NY3_s is composed of a multiplier Multiplier_s and an accumulator Adder_s. Under the control of the clock synchronization signal, the input from the upper layer neuron enters the neuron serially and multiplies its weight, and the multiplication result is accumulated in the accumulator Adder_s.

[0025] The artificial neural network algorithm system uses a 3-2-1 network, and the top module xor_s includes 3 input neurons, 2 hidden layer neurons, and 1 output layer neuron.

[0026] see figure 1 , the whole neural network can be ...

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 present invention provides a hardware implementation method and system of an artificial neural network algorithm, which overcomes the disadvantages of expensive and excessively occupied hardware resources in the existing hardware implementation technology; it includes a top-level module, and the top-level module is composed of several neuron modules The pulse input terminal of the neuron module inputs a given weight product, and its output terminal is fitted to the Sigmoid function of the neural network through a normal distribution random generator and a nonlinear transformer, and then through a pulse converter, converted Pulse output; the hardware implementation method and system of the neural network algorithm of the present invention are compared with existing implementation technology, overcome the embarrassment of completing parallel computing in the serial mode in the software implementation in the past, and replace the disadvantages of software implementation; Due to the optimization of the algorithm, a lot of hardware resources are saved. It is implemented on the FPGA platform, and the cost is low. Using the optimized algorithm, it avoids the dilemma that the Sigmoid function is difficult to implement with hardware, and uses the accumulation method to save hardware resources.

Description

technical field [0001] The invention relates to an artificial neural network algorithm, in particular to a hardware implementation method and system of the artificial neural network algorithm through FPGA hardware. Background technique [0002] Artificial neural network is a kind of imitating human intuitive thinking, and it is a nonlinear dynamic system, which is characterized by distributed storage of information and parallel collaborative processing. Although the structure of a single neuron is extremely simple and its functions are limited, the behaviors that can be realized by a network system composed of a large number of neurons are extremely colorful. The study of artificial neural network is to effectively use this characteristic of the human brain, which belongs to an emerging edge and interdisciplinary science. The study of neural network will definitely have a profound impact on computer science and intelligence science, and improve the intelligence of computers....

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): G06N3/06
Inventor 马立新李长乐张学佳
Owner UNIV OF SHANGHAI FOR SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com