Physics-based high-performance bionic underwater robot simulation training platform construction method

An underwater robot and simulation training technology, applied in neural learning methods, design optimization/simulation, instruments, etc., can solve problems such as limited fluid simulation area and low performance, lack of fluid simulation training environment, and imperfect underwater robot modeling. , to save storage space and improve computing efficiency.

Pending Publication Date: 2022-07-26
SHANGHAI TECH UNIV
View PDF0 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is: to propose a high-performance bionic underwater robot simulation training platform based on physics, to solve the problem of imperfect modeling of underwater robots in the prior art,

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
  • Physics-based high-performance bionic underwater robot simulation training platform construction method
  • Physics-based high-performance bionic underwater robot simulation training platform construction method
  • Physics-based high-performance bionic underwater robot simulation training platform construction method

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach

[0104] In the processing of the simulation domain boundary, the present invention adopts the "unbalanced extrapolation" boundary condition. At the boundary, f is missing towards the interior of the fluid domain due to the lack of external grid points. For these missing f i , the present invention by specifying its non-equilibrium distribution f i neq (ρ w , u w ) approximates the non-equilibrium distribution of adjacent fluid lattice points, where ρ w represents the fluid density at the boundary, u w Represents the fluid velocity at the boundary. The specific implementation of this boundary condition is as follows:

[0105] f i (x b , t)=f i eq (ρ w , u w )+(f i (x f , t)-f i neq (ρ f , u f ))

[0106] where x brefers to the grid point at the boundary of the fluid domain, x f refers to the distance x b The nearest fluid lattice point, t represents the t-th time step, ρ f means x f fluid density at u f means x f fluid velocity at .

[0107] Figure ...

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 physics-based high-performance bionic underwater robot simulation training platform construction method which can be used for training a fish-shaped underwater robot. The platform is composed of a robotic fish modeling module using a joint body with skin, a GPU-based high-performance local bidirectional fluid-solid coupling simulation module (capable of processing a finite field and an infinite field), and a reinforcement learning module. A plurality of bionic robot models are provided, a series of underwater reference control tasks are formulated, and an existing reinforcement learning training method is applied to obtain a corresponding control strategy. According to the method, high-performance fluid simulation requirements of a finite domain and an infinite domain can be processed, a model of any bionic underwater robot can be built, an underwater control task can be customized, and intelligent strategy training, migration and integration can be realized.

Description

technical field [0001] The invention belongs to the fields of computer graphics, fluid simulation, bionic robots and machine learning, and relates to a construction method of a physics-based high-performance bionic underwater robot simulation training platform. Background technique [0002] Bionic underwater robots have superior maneuverability, propulsion efficiency and realistic visual appearance. These advantages have prompted the research, production and application of bionic soft robots and bionic fish-like robots in the past few years. Underwater robots have important applications in many aspects, such as marine education, navigation and rescue, seabed exploration, scientific surveying, etc. However, due to the high cost of hardware, unstable environment, endurance, slow data collection and many other practical problems in studying the control mode and motion mode of robots in a real environment, training robots to obtain motion control strategies that approximate or ...

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
IPC IPC(8): G06F30/17G06F30/23G06F30/25G06F30/27G06F30/28G06N3/04G06N3/08G06F111/08G06F113/08G06F119/14
CPCG06F30/17G06F30/23G06F30/25G06F30/27G06F30/28G06N3/084G06F2111/08G06F2113/08G06F2119/14G06N3/048G06N3/045
Inventor 刘晓培刘文基柏凯
Owner SHANGHAI TECH UNIV
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap