Scholarlink

<p>The crested porcupine optimizer (CPO) is a recently developed metaheuristic algorithm noted for its efficacy in numerical optimization. However, its application in large-scale, high-dimensional configurations—such as multi-UAV swarm coordination—is often hindered by computational bottlenecks, including an imbalance between exploration and exploitation and a lack of cooperative search mechanisms. To address these challenges and meet the stringent real-time performance requirements of complex mission planning, this paper proposes an improved crested porcupine optimizer (ICPO) designed for high-performance computing (HPC) environments. The ICPO incorporates four synergistic enhancements: (1) Dynamic Cosine-based Adaptive Parameter Adjustment (DCAPA): This strategy leverages the nonlinear characteristics of the cosine function to dynamically modulate the search intensity, ensuring an optimal transition from global exploration to local refinement while reducing the redundant computational overhead. (2) Proactive Response Mobility (PRM): Inspired by the “stimulus–response” concept, this mechanism replaces stochastic “blind escape” with an environment-aware adaptive migration, enabling individuals to perceive high-dimensional landscapes from multiple directions—a structure highly conducive to parallel processing and distributed execution. (3) Guided Differential Mutation (GDM): By integrating global best guidance with stochastic perturbations, this strategy accelerates convergence and stabilizes the algorithm against environmental uncertainty, providing the computational efficiency necessary for time-critical decision-making. (4) Enhanced Tactical Movement (ETM): This strategy reinforces cooperative adaptation among swarm members, significantly improving search coordination and the ability to escape complex local optima in resource-intensive optimization tasks. The performance of ICPO was rigorously evaluated using the IEEE CEC 2017, 2019, and 2020 benchmark suites. Results demonstrate that ICPO exhibits superior scalability and stability when solving complex, high-dimensional problems. Furthermore, simulation experiments in high-threat, large-scale mountainous terrains show that ICPO consistently generates energy-efficient paths with rapid convergence and low computational latency. These outcomes demonstrate that ICPO provides an effective and robust solution for real-time multi-UAV path planning, particularly in parallel-computing frameworks characterized by high dimensionality and multi-objective constraints.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An improved crested porcupine optimizer with applications to UAV swarm path planning in complex environments

  • 期刊论文

  • 发表日期：

    2026年4月24日

查看全文
  • Yahong Zhai,
  • Xingtong Hang,
  • Mao Xi,
  • Yifei Zhang,
  • Longyan Xu,
  • Yangbing Zheng

摘要

The crested porcupine optimizer (CPO) is a recently developed metaheuristic algorithm noted for its efficacy in numerical optimization. However, its application in large-scale, high-dimensional configurations—such as multi-UAV swarm coordination—is often hindered by computational bottlenecks, including an imbalance between exploration and exploitation and a lack of cooperative search mechanisms. To address these challenges and meet the stringent real-time performance requirements of complex mission planning, this paper proposes an improved crested porcupine optimizer (ICPO) designed for high-performance computing (HPC) environments. The ICPO incorporates four synergistic enhancements: (1) Dynamic Cosine-based Adaptive Parameter Adjustment (DCAPA): This strategy leverages the nonlinear characteristics of the cosine function to dynamically modulate the search intensity, ensuring an optimal transition from global exploration to local refinement while reducing the redundant computational overhead. (2) Proactive Response Mobility (PRM): Inspired by the “stimulus–response” concept, this mechanism replaces stochastic “blind escape” with an environment-aware adaptive migration, enabling individuals to perceive high-dimensional landscapes from multiple directions—a structure highly conducive to parallel processing and distributed execution. (3) Guided Differential Mutation (GDM): By integrating global best guidance with stochastic perturbations, this strategy accelerates convergence and stabilizes the algorithm against environmental uncertainty, providing the computational efficiency necessary for time-critical decision-making. (4) Enhanced Tactical Movement (ETM): This strategy reinforces cooperative adaptation among swarm members, significantly improving search coordination and the ability to escape complex local optima in resource-intensive optimization tasks. The performance of ICPO was rigorously evaluated using the IEEE CEC 2017, 2019, and 2020 benchmark suites. Results demonstrate that ICPO exhibits superior scalability and stability when solving complex, high-dimensional problems. Furthermore, simulation experiments in high-threat, large-scale mountainous terrains show that ICPO consistently generates energy-efficient paths with rapid convergence and low computational latency. These outcomes demonstrate that ICPO provides an effective and robust solution for real-time multi-UAV path planning, particularly in parallel-computing frameworks characterized by high dimensionality and multi-objective constraints.