Abstract:The distributed telescope array offers promise for conducting large-sky-area, high-frequency time domain surveys. Multiple telescopes can be deployed at each observation site, so intra-site observation task scheduling is crucial for enhancing observation efficiency and quality. Efficient use of observable time and rapid response to special situations are critical to maximize scientific discovery in time domain surveys. Besides, the competing scientific priorities, time-varying observation conditions, and capabilities of observation equipment, lead to a vast search space of the scheduling. So with the increasing number of telescopes and observation fields, balancing computational time with solution quality in observation scheduling poses a significant challenge. Informed by the seminal contributions of earlier studies on a multilevel scheduling model and global scheduler for time domain telescope array, this study is devoted to further exploring the site scheduler. Formulating the observation scheduling of multiple telescopes at the site as a cooperative decision-making problem, this paper proposes GRRIS, a real-time intra-site observation scheduling scheme for telescope array using graph and reinforcement learning. It employs a graph neural network to learn node features that can embed the spatial structure of the observation scheduling. An algorithm based on multi-agent reinforcement learning is designed to efficiently learn the optimum allocation policy of telescope agents to field nodes. Through numerical simulations with real-world scenarios, GRRIS can achieve up to a 22% solution improvement over the most competitive scheme. It offers better scalability and sub-second decision speed, meeting the needs of observation scheduling control for future distributed telescope arrays.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to efficiently arrange observation tasks to improve the efficiency and quality of observation when the distributed telescope array conducts large - sky - area and high - frequency time - domain sky surveys. Specifically: 1. **Multi - telescope Collaboration**: Multiple telescopes are deployed within each observatory station. Therefore, it is necessary to optimize the collaborative scheduling among these telescopes to enhance the efficiency and quality of observation. 2. **Effective Utilization of Observable Time**: Quickly respond to the observation requirements in special cases (such as the appearance of transient objects) and maximize the opportunities for scientific discovery. 3. **Complex Scheduling Challenges**: As the number of telescopes and the number of observation fields increase, it becomes very challenging to balance the computing time and the quality of solutions. This involves factors such as competing scientific priorities, changing observation conditions, and the capabilities of observation equipment, resulting in a huge scheduling search space. To solve these problems, the paper proposes GRRIS (Graph - based and Reinforcement - learning - based Real - time Intra - station Observation Scheduling Scheme). The main contributions of GRRIS include: 1. **Efficient Intra - station Observation Scheduling Strategy**: Through the modeling of cooperative decision - making problems, an intra - station observation scheduling strategy for distributed multi - telescope observation systems is proposed, aiming to maximize the total observation quality and minimize the computing time. 2. **Graph - based Reinforcement - learning Scheme**: Transform the complex non - convex optimization problem into a sequential decision - making problem, and design an algorithm based on graph neural networks and multi - agent reinforcement learning to learn the optimal observation scheduling strategy. 3. **Superior Performance**: Through simulation experiments with actual data, GRRIS is superior to five other representative scheduling optimization methods in terms of decision - making speed, robustness to changes in observation time and geographical distribution, and scalability. In addition, the paper also shows that GRRIS can effectively schedule between different numbers of observation tasks without retraining the model, further proving its application potential in large - scale distributed telescope arrays. ### Formula Summary 1. **Observation Cost Calculation**: \[ L_i=\sum_{j = 1}^{T_i}c(t_j,p_a,p_v) \] where \(c(t_j,p_a,p_v)\) represents the observation cost of telescope \(a\) for the observation field \(v\) at time \(t_j\), and \(p_a\) and \(p_v\) represent the positions of the telescope and the observation field respectively. 2. **Objective Function (Min - Max Optimization)**: \[ \min\max_{i}L_i \] that is, minimize the largest observation cost among all telescope agents. 3. **Node Embedding Update**: \[ f_v^{(k + 1)}=\text{ReLU}(\theta_1f_v^{(k)}+\theta_2\max_{u\in\mathcal{N}(v)}h(e_{uv},f_u^{(k)})) \] where \(\mathcal{N}(v)\) represents the set of neighbors of node \(v\), and \(h(e_{uv},f_u^{(k)})\) is the attention vector predicted by the edge type. 4. **Attention Mechanism Calculation**: \[ w_{ij}=\frac{\exp(u_{ij})}{\sum_{j'= 1}^n\exp(u_{ij'})} \] where \(u_{ij}\) represents the compatibility between the query vector and the node feature. Through these formulas and methods, GRRIS realizes efficient multi - telescope collaborative observation scheduling and significantly improves the efficiency and quality of observation.

GRRIS: a real-time intra-site observation scheduling scheme for distributed survey telescope arrays

GRRIS: A Real-time Intrasite Observation Scheduling Scheme for Distributed Survey Telescope Arrays

Deep Reinforcement Learning-Based Periodic Earth Observation Scheduling for Agile Satellite Constellation.

A Multilevel Scheduling Framework for Distributed Time-domain Large-area Sky Survey Telescope Array

Relay Observation Scheduling of Global Distributed Telescope Array Based on Integer Programming

Scheduling Remote Access to Scientific Instruments in Cyberinfrastructure for Education and Research

Time-dependent short-term observational scheduling method for Yunnan 40 m Radio Telescope using a genetic algorithm

Multi-satellites Imaging Scheduling Using Individual Reconfiguration Based Integer Coding Genetic Algorithm

Self-Driving Telescopes: Autonomous Scheduling of Astronomical Observation Campaigns with Offline Reinforcement Learning

An onboard periodic rescheduling algorithm for satellite observation scheduling problem with common dynamic tasks

A Hierarchical Resource Scheduling Method for Satellite Control System Based on Deep Reinforcement Learning

A DRL-based online real-time task scheduling method with ISSA strategy

Earth Observation Satellite Scheduling with Graph Neural Networks

A Data-Driven Parallel Scheduling Approach for Multiple Agile Earth Observation Satellites

Event-Triggered Deep Reinforcement Learning for Dynamic Task Scheduling in Multi-Satellite Resource Allocation

Deep Reinforcement Learning for the Agile Earth Observation Satellite Scheduling Problem

A Heuristic Method for Sensor Scheduling in Infrared LEO Constellation.

Asymmetric path-relinking based heuristics for large-scale job scheduling problem in TDRSS

Multi-Task Multi-Agent Reinforcement Learning for Real-Time Scheduling of a Dual-Resource Flexible Job Shop with Robots

Learning-based constellation scheduling for time-sensitive space multi-target collaborative observation

Coordinated Scheduling of Air and Space Observation Resources Via Divide and Conquer Framework and Iterative Optimization