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

Robot path planning system based on memristive cross array and q-learning

A cross-array and path planning technology, applied in the field of memristive cross-array and reinforcement learning, can solve problems such as slow convergence speed and long machine learning time, and achieve the effects of improving flexibility, avoiding state explosion problems, and reducing learning time

Active Publication Date: 2019-07-26
SOUTHWEST UNIV
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] However, after continuing research, it has been found that the existing Q-learning system based on the memristive cross-array has the following defects: the convergence speed is too slow, and the machine learning time is relatively long

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
  • Robot path planning system based on memristive cross array and q-learning
  • Robot path planning system based on memristive cross array and q-learning
  • Robot path planning system based on memristive cross array and q-learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0039] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail in conjunction with the accompanying drawings and specific embodiments, and the description here does not mean that all the subjects corresponding to the specific examples stated in the embodiments are in cited in the claims.

[0040] Compared with the prior patent application 201210188573.2, the technical solution disclosed in the present invention mainly proposes two improvements;

[0041] (1) On the basis of improved Q-learning (Q learning), introduce and combine the memristive cross array to store the Q value;

[0042] (2) Based on the improved Q-learning and memristive cross array, the robot path planning is realized.

[0043] Specifically, the robot perceives the current state s through the environmentt ∈S (S represents the set of all states), and execute the corresponding action a t ∈A (A represents...

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 discloses a robot path planning system based on a memristive crossbar array and Q learning. On one hand, a returned report of a distance target point in the Q learning is expressed by a function rather than a discrete point, and on the other hand, when a robot is away from an obstacle for a certain range, the robot is allowed to get close to the target point linearly. The invention provides a memristive crossbar array model with continuous input / output, dynamic variable resistance and the nano size, and voltage and imposing time required for changing the memristive value are specifically derived. At last, through experimental analysis, the validity of the scheme is verified. The invention also provides a novel scheme for achieving robot path planning system by use of a memristive crossbar array, so new ideas are provided for quite wide application of memristor and robot route planning.

Description

technical field [0001] The invention relates to a memristive cross array and reinforcement learning technology, in particular to a robot path planning system based on a memristive cross array and Q-learning. Background technique [0002] Reinforcement learning is a well-known unsupervised machine intelligence learning algorithm, and it is widely used in artificial intelligence and other fields. Famous reinforcement learning algorithms include: TD algorithm proposed by Sutton in 1988; R-Learning algorithm proposed by Schwartz; Q-Learning algorithm proposed by Watkins in 1989; Q(λ) algorithm proposed by Peng and Williams in 1996, etc. Among them, the relatively important Q-Learning algorithm is widely used in robot path planning. [0003] However, the traditional Q-Learning uses the lookup table method to store the generated Q value, so when the state space gradually increases, the state explosion may occur due to the huge storage space required by the Q-Learning algorithm, m...

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): G05D1/02G06N20/00
CPCG05D1/0221G06N20/00
Inventor 胡小方马异峰段书凯贾鹏飞彭小燕
Owner SOUTHWEST UNIV
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