Skip to main navigation Skip to search Skip to main content

Data-driven-based Nonlinear Optimal Guidance with Impact Time and Angle Constraints

  • Xuwei Quan
  • , Tao Song
  • , Hong Tao*
  • , Denghui Dou
  • *Corresponding author for this work
  • Beijing Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper proposes a data-driven nonlinear optimal guidance method with impact time and angle control. First, the optimality conditions of the nonlinear model are derived based on the nondimensionalized motion model. Then, A parameterized reverse dynamic model is constructed to generate optimal interception trajectories. Based on this model, a feedforward neural network is trained to capture the functional mapping from relative motion states to the optimal commands, enabling real-time guidance. The simulation results verify that the proposed method achieves interception with optimal energy consumption, subject to strict impact time and angle constraints.

Original languageEnglish
Title of host publication2025 2nd International Conference on Unmanned Systems and Automation Control, ICUSAC 2025
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages45-52
Number of pages8
ISBN (Electronic)9798331569037
DOIs
Publication statusPublished - 2025
Externally publishedYes
Event2nd International Conference on Unmanned Systems and Automation Control, ICUSAC 2025 - Changsha, China
Duration: 26 Dec 202528 Dec 2025

Publication series

Name2025 2nd International Conference on Unmanned Systems and Automation Control, ICUSAC 2025

Conference

Conference2nd International Conference on Unmanned Systems and Automation Control, ICUSAC 2025
Country/TerritoryChina
CityChangsha
Period26/12/2528/12/25

Keywords

  • Data-driven
  • Impact Time and Angle Control
  • Neural network
  • Nonlinear optimal guidance

Fingerprint

Dive into the research topics of 'Data-driven-based Nonlinear Optimal Guidance with Impact Time and Angle Constraints'. Together they form a unique fingerprint.

Cite this