低空巡检技术以其在降本增效方面的显著优势，正逐步发展成为现代巡检体系中的关键技术。实现高效的无人机巡检作业，依赖于一个完善的顶层规划系统，其核心问题可归纳为机巢选址、任务分配与路径规划这三个紧密的决策层。本文系统性地综述该集成规划体系的研究现状与建模方法。首先，本文从巡检对象、场景与组合方式等维度对任务特性进行了分类剖析；在此基础上，重点阐述了机巢选址、任务分配与路径规划这三个核心层级的建模方法、优化目标与约束体系。通过“选址-分配-路径”三层研究框架的综述，为该领域提供了更系统、更具层级化的分析视角。最后，针对当前研究在系统协同、实时性与环境适应性方面的瓶颈，本文展望了未来发展方向，包括三级联合优化、云-边-端协同计算、数据驱动与物理模型融合的鲁棒规划等。本文旨在为无人机巡检智能规划系统的理论研究与工程实践提供系统性参考。

Low-altitude inspection technology, noted for its remarkable advantages in cost reduction and efficiency improvement, is increasingly becoming a key enabling technology in modern inspection systems. The efficient operation of unmanned aerial vehicle (UAV) inspections depends on a well-structured top-level planning framework, the core of which comprises three closely interrelated decision-making layers: site selection, task assignment, and path planning. This paper presents a systematic review of current research and modeling methodologies concerning this integrated planning framework. Firstly, task characteristics are categorized and analyzed from multiple dimensions, including inspection targets, operational scenarios, and combination modes. Building on this foundation, the modeling approaches, optimization objectives, and constraint systems of the three core layers site selection, task assignment, and path planning are discussed in detail. By establishing a three-tier analytical framework, the review provides a more systematic and hierarchical perspective for the field. Finally, in response to existing research challenges in system coordination, real-time adaptability, and environmental robustness, this paper highlights several future research directions, including three-level integrated optimization, cloud–edge–device collaborative computing, and robust planning that integrates data-driven and physics-based models. The goal of this study is to offer a comprehensive reference for both theoretical research and engineering practice in the advancement of intelligent planning systems for UAV inspection.