Performance enhancement of homing missiles in air defense systems through model predictive control integrated with neural networks.

Document Type : Dynamics, Vibrations, and Control

Authors

1 PhD Student, Khajeh Nasir University, Tehran, Iran

2 Assistant Professor, Imam Ali Officer University, Tehran, Iran

Abstract

Achieving high precision and optimal performance in missile guidance systems necessitates a unified design framework with simultaneous optimization of multiple system components and constraints. This paper presents an innovative and effective approach to improving the overall performance of missile guidance by employing a Model Predictive Controller (MPC) integrated with a neural network-based identifier. The proposed method focuses on accurate three-dimensional modeling and real-time trajectory optimization between the missile and its maneuvering target, aiming to minimize guidance errors and reduce total flight time. By leveraging a neural network to learn the nonlinear dynamics of the system, the controller eliminates the need for an exact analytical model, significantly decreasing dependency on traditional modeling techniques. This not only enhances modeling flexibility and adaptability but also contributes to lower development and implementation costs in practical scenarios. Simulation results confirm that the proposed MPC-based controller significantly outperforms conventional strategies such as PID in terms of miss distance reduction and faster interception under dynamic conditions.

Keywords

Main Subjects

Files

History

  • Receive Date: 09 October 2025
  • Revise Date: 13 November 2025
  • Accept Date: 29 November 2025
  • Publish Date: 22 December 2025

Share

How to cite

Statistics