Abstract
The Multi-channel Antarctic Solar Telescope (MARST) is a pioneering instrument designed for simultaneous multi-wavelength solar observations from the South Pole. This paper describes the development of a robust, three-layered software control system that manages the telescope's dual-tube configuration—one for Photosphere observation (utilizing 11 filters) and one for Chromosphere observation (Hα waveband). By integrating the Experimental Physics and Industrial Control System (EPICS) with the Remote Telescope System (RTS2), the architecture achieves the high levels of autonomy, hardware abstraction, and tracking precision required for Earth's most extreme environments.
Introduction
Operating astronomical instrumentation in Antarctica presents significant engineering hurdles, including extreme temperatures, minimal human intervention, and the need for high-precision tracking over long durations. MARST, China's first solar telescope destined for this region, requires a control system capable of coordinating diverse hardware while maintaining scientific accuracy. The telescope features two optical paths that must be synchronized to capture solar data across different atmospheric layers simultaneously, necessitating a sophisticated software brain to manage resources and scheduling.
System Architecture
The MARST software employs a modular, three-layer architecture designed for stability and flexibility, running primarily on a Linux environment.
Device Control Layer (EPICS)
At the lowest level, the system utilizes the Experimental Physics and Industrial Control System (EPICS). EPICS provides a distributed control framework that abstracts hardware interactions into "Process Variables" (PVs). Dedicated Input/Output Controllers (IOCs) were developed for the telescope mount, focusers, filter wheels, Hα filters, dome, and CCD cameras. A notable engineering solution involved the PI CCD camera, which required a legacy Linux kernel for its drivers. To maintain system integrity, the authors used QEMU virtualization to host a CentOS 6 environment, bridging it to the modern EPICS stack via a virtualized network.
Observation Operation Layer (RTS2)
The middle layer utilizes the Remote Telescope System (Version 2) to manage autonomous scheduling. A custom RTS2 executor module was developed to map EPICS properties into high-level observation plans via XML configurations. This layer handles the complex logic of coordinating the two telescope tubes, ensuring that while both can capture images, only one controls the shared mount at any given time.
Autonomous Coordination and Self-Guiding
To maximize scientific output without manual oversight, the system implements priority-based execution. Different observation plans—including normal, dark, flat-field, and bias modes—are assigned priorities, allowing the executor to automatically resolve resource conflicts and manage the mount and camera resources effectively.
A critical feature of the MARST software is its image-based self-guiding module. Due to inherent mechanical drift in telescope mounts over long observation periods, the software performs real-time luminosity centroid calculations on every solar image. If the calculated offset exceeds a predefined threshold, the system automatically applies a correction to the mount. This software-driven approach effectively compensates for hardware limitations, ensuring the Sun remains perfectly centered throughout the observation cycle.
User Interface and Results
For remote monitoring and manual intervention, a frontend GUI was developed using PyQt5 and QML. This interface connects to the underlying system via rts2-httpd and rts2-proxy modules, providing a comprehensive view of device status, system logs, and live imagery.
The system has undergone extensive testing at the Nanjing Institute of Astronomical Optics and Technology. Results indicate that the EPICS-RTS2 integration is highly robust, successfully managing simultaneous dual-path observations and maintaining stability over long-term operation. All EPICS IOCs functioned properly under heavy testing loads, demonstrating the system's readiness for deployment.
Conclusion
The MARST control software provides a scalable and autonomous solution for solar observation in extreme environments. By combining the hardware-level reliability of EPICS with the high-level scheduling capabilities of RTS2, the team has created a system ready for the challenges of the Antarctic. Future work will focus on refining autonomous error-handling routines to address specific mechanical failures associated with sub-zero conditions, such as sensor icing or frozen gears.
Summary updated