系统工程与电子技术 ›› 2026, Vol. 48 ›› Issue (8): 2648-2658.doi: 10.12305/j.issn.1001-506X.2026.08.13

• 系统工程 • 上一篇    

基于MCP的无人集群作战效能评估方法

苗延飞1, 陈海昕1, 高原2, 张旭3, 蔡清2   

  1. 1. 清华大学航天航空学院,北京 100084
    2. 西安电子科技大学电子工程学院,陕西 西安 710071
    3. 西安电子科技大学人工智能学院,陕西 西安 710071
  • 收稿日期:2025-08-11 修回日期:2025-10-01 出版日期:2026-07-12 发布日期:2026-07-12
  • 通讯作者: 蔡清
  • 作者简介:苗延飞(1988—),男,高级工程师,博士,主要研究方向为无人机总体设计、体系作战试验、无人集群作战智能决策
    陈海昕(1974—),男,教授,博士,主要研究方向为飞行器设计
    高 原(1996—),男,副教授,博士,主要研究方向为基于联邦学习的隐私保护
    张 旭(1992—),男,副教授,博士,主要研究方向为多智能体强化学习理论方法及应用
  • 基金资助:
    中央高校基本科研业务费专项资金(ZYTS25137);陕西省自然科学基础研究计划(2025JC-YBMS-660)资助课题

Combat effectiveness evaluation method of unmanned swarm based on MCP

Yanfei MIAO1, Haixin CHEN1, yuan GAO2, Xu ZHANG3, Qing CAI2   

  1. 1. School of Aerospace Engineering,Tsinghua University,Beijing 100084,China
    2. School of Electronic Engineering,Xidian University,Xi’an 710071,China
    3. School of Artificial Intelligence,Xidian University,Xi’an 710071,China
  • Received:2025-08-11 Revised:2025-10-01 Online:2026-07-12 Published:2026-07-12
  • Contact: Qing CAI

摘要:

为解决当前无人集群作战效能评估研究普遍依赖专家经验和预设规则而导致难以满足实战化评估的智能化与自动化需求问题,提出一种基于模型上下文协议的无人集群作战效能评估方法。该方法构建了评估指标的统一数学建模框架,结合自然语言解析、模型调度与数据处理组件,支持用户以文本形式输入评估需求,由智能体自动完成评估模型选择、数据处理流程配置与评估计算任务,最终生成多维度评估结果。通过设计典型作战场景进行实验验证,结果表明所提方法在评估精度、响应效率和交互便捷性方面具有显著优势,能够有效降低人工干预强度,提高评估的智能化水平。研究成果可为无人集群作战效能智能评估系统的构建提供方法支撑与技术参考。

关键词: 无人集群作战, 作战效能评估, 模型上下文协议, 人工智能智能体

Abstract:

To address the issue that current research on unmanned swarm combat effectiveness evaluation generally relies on expert experience and preset rules, making it difficult to meet the intelligent and automated requirements of practical combat evaluation, an unmanned swarm combat effectiveness evaluation method based on model-context protocol is proposed. The method establishes a unified mathematical modeling framework for evaluation indicators and integrates components for natural language parsing, model scheduling, and data processing, supports users inputting evaluation requirements in text format. The agent autonomously selects evaluation models, configures data processing pipelines, and performs assessment tasks, ultimately generating multidimensional evaluation results. Experiments based on typical combat scenarios demonstrate that the proposed method has significantly advantages in evaluation accuracy, response efficiency, and interaction convenience. It effectively reduces human intervention and enhances the level of intelligent. The findings provide methodological and technical reference for building intelligent evaluation system for unmanned swarm combat effectiveness.

Key words: unmanned swarm combat, combat effectiveness evaluation, model-context protocol(MCP), artificial intelligence(AI) agent

中图分类号: