Earth Observation Satellites

The past twentieth century is an extraordinary era with significance historical revolutions and great progress in science. For the first time, the human beings made a step out of the earth, where they lived and raised for hundreds of thousands of years, and made a step into the magnificent “Space age.” From the first artificial satellite “Sputnik 1” launched by the Soviet Union in 1957 to the early twenty-first century, aerospace technology has achieved significant progress in the transition from the laboratory simulation to the practical application. Meanwhile, with the developments of the spacecraft design, the applications of the aerospace technology have permeated to different societies including scientific research, economical events, and security operations.

In order to analyze the essence of the EOSs task scheduling problem, this chapter first describes the working process and mechanism of EOSs and data transmission resources. Then, the elements of the EOSs resource scheduling problem, including the elements of earth observation tasks, satellite resources, data transmission resources, and the elements of the optimization objective function, are clarified. On this basis, the main characteristics and challenges of the EOSs task scheduling problem are analyzed, and the differences and connections between the EOSs task scheduling problem and the classical task scheduling problem are discussed.

Ground-based centralized satellite task scheduling is a well-studied problem in the field of satellite operations. This method assumes that a single operation and control center manages all satellite resources and establishes a mathematical model for earth observation satellites (EOSs) and their resource scheduling. A centralized optimization algorithm is then used to solve the problem. The ground-based centralized satellite task scheduling problem can be divided into two main subproblems: the earth observation task scheduling problem and the satellite data transmission scheduling problem. The former involves determining when the satellite should use its sensors to observe which targets, while the latter involves deciding when the satellite should transmit observation data to which ground station. The two problems have essential relationship due to constraints on the use of satellite sensor and resource constraints such as satellite energy and onboard memory storage.

Chapter 3 introduced a ground-based centralized satellite task scheduling model and methods that can be applied to the task scheduling problem of earth observation satellites (EOSs) in static scenarios. The static scenarios refer to situations in which the satellite resources and observation tasks engaged in scheduling scenarios will not change once the scheduling process begins. However, EOSs operate in a complex environment full of changes. They may experience temporary failures, be repaired, or have new observation tasks submitted at any time. Failing to account for these changes during the task scheduling process will inevitably result in suboptimal scheduling outcomes.

This chapter presents an introduction to satellite task scheduling methods in distributed scenarios. Both distributed and centralized satellite task scheduling have their own advantages and disadvantages, which are applicable to different application scenarios. Together, they form the mainstream technology system of multi-satellite joint task scheduling. The distributed satellite task scheduling is based on the multi-agent system (MAS) theory in distributed artificial intelligence. In this approach, each satellite participating in the scheduling process is considered an intelligent agent undertaking earth observation tasks based on its capabilities and benefits. These satellite agents interact with each other to rapidly form a multi-satellite observation programme. The chapter first describes the multi-satellite distributed task scheduling problem and analyzes its application scenarios along with the multi-agent system architecture. Then, it establishes the distributed satellite task scheduling model. Finally, it proposes the distributed satellite task scheduling method based on the cooperation mechanism, such as contract network protocol and blackboard model.

Recent advancements in artificial intelligence and space science technology have led to the emergence of a new generation of satellites with onboard data processing and autonomous task scheduling capabilities. Such a satellite is no longer a passive executor, but rather a decision-maker that can actively generate or modify observation programme based on changes in the space environment, equipment status, and available resources. Its capability to respond quickly to uncertain space environments and unexpected satellite states surpasses that of traditional satellites. Due to the limitation of the onboard processor’s computing power, the task scheduling algorithms used in ground computing devices cannot be applied onboard. Therefore, new methods and mechanisms must be investigated to adapt to these changes. This chapter introduces some mainstream satellite onboard autonomous task scheduling models and techniques.

In Chaps. 3–6, some typical satellite task scheduling methods for centralized EOS task scheduling, dynamic rescheduling, distributed scheduling, and onboard autonomous scheduling are described in detail, which will eventually be applied to the satellite task planning and scheduling system. This chapter introduces several typical satellite task scheduling systems and simulation tools involved in the practice and focuses on analyzing more typical distributed task scheduling systems.

With the rapid development of the global aerospace industry, the number and types of satellites in orbit and supporting resources are increasing, resulting in unprecedented attention to EOSs task scheduling in both academic and industrial circles. As a result, various new techniques and methods are emerging in this field. One of the main objectives of this book is to organize and present the current existing technologies and to explore future technological trends in this area.