Intelligent agent behavior interpretation method based on causal relationship inference

A technology of causality and agents, applied in biological neural network models, instruments, computing models, etc., can solve problems such as the difficulty of explaining agent behavior models, and achieve the effect of optimizing intelligence and concise behavior characteristics

Pending Publication Date: 2022-04-15
沈阳飞机设计研究所扬州协同创新研究院有限公司
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve or improve the defects that the above-mentioned agent behavior model is difficult to explain

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
  • Intelligent agent behavior interpretation method based on causal relationship inference
  • Intelligent agent behavior interpretation method based on causal relationship inference
  • Intelligent agent behavior interpretation method based on causal relationship inference

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] The technical solutions of the present invention will be further described below with reference to the accompanying drawings and examples.

[0041] Causal structure learning is conducted by keeping the data during the aircraft's intelligent training process. The main data of the data includes the current time-only, the relative distance (Distance), a closed rate, red, and blue, and a velocity, a speed (Velocity), and climb. H_DOT), angle (alpha), side slip angle (Beta), overload (N_LOAD), blood volume (Blood), remaining oil volume (OIL).

[0042] Based on independent testing, the independence is judged by the calculation of sample correlation coefficient between two or two variables, combined with Markov assumptions verifying causality on the basis of independence. For a relatively complex data such as the relative coordinates of the aircraft, the method of adding noise on the model, constructing the causal structure diagram between the data, using multi-layer perception ma...

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 an agent behavior interpretation method based on causal relationship inference, and belongs to the technical field of auxiliary decision making and causal inference. The method comprises the following steps of: performing training data acquisition on an intelligent agent trained by adopting reinforcement learning, wherein the data comprises an environment state, an adopted action and reward information in an intelligent agent training process; performing offline training on the data through causal relationship discovery and data regression fitting methods, and outputting a reasonable behavior causal relationship model; and performing online interpretation on the behaviors of the intelligent agent by using the behavior causal relationship model. According to the method, a good behavior interpretation effect can be achieved.

Description

Technical field [0001] The present invention is assisted in decision-making, and in fact, in particular, in particular, the present invention relates to a method of strengthening the interpretation of learning intelligent behavior. Background technique [0002] In recent years, the value of drones has been highlighted in several local wars, and has received more and more researchers. UAV air confronted the game decision-making, directly reflecting its intelligence level, affecting its combat effectiveness when compliant with someone or drones. Currently, the mainstream method of unmanned airborne counterfeiting decision-making issues is to strengthen learning, such as policy gradient, actor-critic, etc. Strengthening learning does not require huge training data sets and adequate priori knowledge, and smart body can learn from zero, constantly learning, adjusting its own strategy, and achieves the choice of optimal behavior during the continuous interaction with the environment. H...

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 Applications(China)
IPC IPC(8): G06N3/04G06N20/00G05D1/08G05D1/10
Inventor 王汉朴海音陈永红陶晓洋于津郝一行彭宣淇韩玥杨晟琦叶超樊松源孙阳
Owner 沈阳飞机设计研究所扬州协同创新研究院有限公司
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