摘要
Owing to the escalating demand for earth observation, a solitary satellite will be required to undertake an expanded array of missions, thereby rendering the scheduling of multipoint earth observation satellite imaging missions increasingly intricate. Herein, we propose a proximal policy optimization (PPO) algorithm based on a spiking neural network (SNN) to solve the multipoint satellite mission scheduling problem. Initially, we preprocess the mission–transition time by incorporating satellite attitude constraints and conceptualize the mission planning process as a Markov decision. Then, our methodology integrates SNN with PPO to effectively handle a high-dimensional state space by leveraging temporal information–processing capabilities. The SNN-based actor-critic network is trained to enhance the scheduling policy via PPO. Our method exhibits superior performance across various satellite orbits, satellite attitude maneuver speeds, and mission scales. In comparison with heuristic methods and traditional reinforcement learning techniques, our method shows a swifter convergence rate and an increased success rate of observation, coupled with superior convergence speed, robustness, and stability. © 2025
| 源语言 | 英语 |
|---|---|
| 期刊论文编号 | 101867 |
| 页数 | 22 |
| 期刊 | Swarm and Evolutionary Computation |
| 卷 | 94 |
| 早期在线日期 | 2月 2025 |
| DOI | |
| 出版状态 | 已出版 - 4月 2025 |
基金资助
National Natural Science Foundation of China (NSFC) [42171341];
关键词
- Satellite mission scheduling
- Reinforcement learning
- Markov decision process
- Proximal policy optimization
收录
- EI
- SCIE
- Scopus
学术指纹
探究 'A spiking neural network based proximal policy optimization method for multi-point imaging mission scheduling of earth observation satellite' 的科研主题。它们共同构成独一无二的学术指纹。引用此
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver