Abstract: Embedded controllers are the core of various industrial control systems and are a typical embedded system. This paper proposes a design method for the embedded controller of an embedded remote monitoring system. Combining embedded system design theory, this paper explains how to design an embedded controller using UML-RT and Platform methods, and provides detailed steps and specific content for the embedded controller system design. Keywords: Embedded; Remote control; UML; Modeling 1 Introduction The core of a remote monitoring system based on embedded instruments is the embedded controller, which is a typical embedded system. The system design of the embedded controller follows the relevant theories of embedded system design. Embedded system design differs from traditional design. Embedded system design needs to consider software response speed, memory space usage, power consumption, etc. System design needs to consider the software and hardware structure as a unified and coordinated whole. Embedded systems tend to use high-platform architecture and object-oriented UML-RT for system analysis and design. Therefore, when designing the embedded controller, we use UML-RT and Platform methods for system design, giving the embedded controller a flexible architecture that allows for convenient and rapid changes or expansion of functions to meet system requirements. 2 UML-RT and Platform 2.1 UML-RT and Embedded System Design UML (Unified Modeling Language) is an object-oriented modeling language proposed by Grady Booch, James Rumbaugh, and Ivan Jacobson, and adopted as an industry standard by the OMG (Object Management Group) in 1997. UML uses a standard notation system to describe models, which can be used to describe the structure and static characteristics, as well as the behavior and dynamic characteristics of a system. It models the system architecture from different perspectives, forming different views of the system. As a general-purpose modeling language, UML provides excellent extension mechanisms, such as stereotypes, tagged values, and constraints, allowing UML to model specific application domains. Utilizing these extension mechanisms, UML-RT, specifically designed for embedded systems, has three basic structures: capsules, ports, and connectors. 2.2 Platform and Embedded System Design We describe embedded systems at an abstract level and provide platform support for this description, recognizing the structure of embedded software and hardware as a unified and coordinated whole. Platform-based design principles have been used in PC design for many years, and in recent years, this approach has become more widespread and formalized, applicable to embedded system design. Using a platform for embedded system modeling addresses the following design challenges: 1. QoS requirements of the application. 2. Resource groups, APIs, and services provided by QoS. 3. Relationships between different abstraction layers, especially resource mapping and usage. In the embedded hardware and software design process, platforms are defined for key connection points. Each platform represents a layer in the design process. Quality of Service (QoS) parameters, such as CPU processing speed and I10 bandwidth, must be defined for the relevant platform. Ideally, the performance parameters of the platform providing QoS should at least meet the performance parameters of the platform requiring QoS. A basic idea of ​​platform-based design is that the platform can support similar or related applications, some of which may appear after the platform is developed. To support these unforeseen programs, the platform services must be scalable. 3 Requirements Analysis of Embedded Controller We use UML use case diagrams to analyze the requirements of the embedded controller. Use case diagrams describe the functions of the system hardware and software from the user's perspective. Based on the functional description of the embedded controller in Chapter 2, we derive the system's use case diagram, as shown in Figure 1. This includes seven different use cases and two different roles. [align=center] Figure 1 Use Case Diagram of Embedded Controller[/align] The user role can view data, set data, and turn the controlled object (smart meter) on and off. The controlled object can receive and send data, and correspondingly open and close operation commands. The data processing center is responsible for data encapsulation, transmission, calculation, and other data processing tasks. 4. Platform Partitioning and Design After completing the system requirements analysis, the system architecture of this platform was partitioned. The platform's system architecture refers to the definition of interfaces between different levels and the functional allocation between upper and lower layers. Each layer has its own system structure. The multi-level hierarchical structure was mainly divided according to functional and logical order. The system architecture of the embedded controller platform is shown in Figure 2. We divide the embedded controller system into three abstract layers: the ARC platform, the application programming interface platform (API), and the system architecture platform. The ARC platform includes a customized application platform (ASP). The ARC platform encompasses the microprocessor storage system, interface circuits, I/O channels, and internal connections between chips. The microprocessor system consists of a family of "microarchitectures," including the Rabbit2000 microprocessor, FLASH, SRAM, and the RTL8019AS network controller chip. The API platform includes the RTOS (Real-Time Operating System), embedded web servers, and device drivers. The API platform is a software abstraction layer that encapsulates the implementation details of the ARC platform. The API platform provides interface services and groups these interfaces by category. For example, users need to know whether the RTOS provides a preemptive task scheduling interface, or whether the embedded web server provides HTTP and TCP/IP interfaces. The ASP platform provides a direct user interface and specialized application services, such as allowing users to view and set control parameters for smart meters through a browser. [align=center] Figure 2 Platform Division[/align] Each platform has corresponding Quality of Service (QoS) requirements. The ARC platform has requirements for power consumption, storage capacity, processing speed, and communication capabilities. Chips such as Rabbit2000 and RTL8019AS can well meet these requirements. The API platform also has requirements for the number of running tasks and task switching time; embedded web servers and RTOS—µc/OS can meet these requirements. The ASP platform requires real-time parameter viewing and setting. Lava Applets can do this very well. Dividing the embedded controller system into platforms and proposing corresponding QoS parameters for each platform allows us to have a global perspective on the entire design from the initial design stage. Because each platform is relatively independent, it has good adaptability to new design requirements and modifications that may arise later. 5. Embedded Controller Design Based on UML-RT After dividing the embedded controller into platforms, we can use the UML-RT notation system to model the embedded controller system. We divide the modeling process into two stages. The first stage considers the system's external interfaces. The second stage considers the internal platform modeling of the system. Based on the requirements analysis in the previous use case diagram, we find that the system needs to interact with the environment through two ports: one is communication with the controlled object, i.e., the smart meter, and the other is interaction with the user. To support user interaction and communication with the controlled object, we define two containers: u: User Interface and t: Transfer Data. The container p: Process Data is used to process data. Containers can contain related software and hardware structures. We further refine t: Transfer Data, as shown in Figure 3. [align=center] Figure 3 Structure of the Transfer Data container[/align] As shown in Figure 3, data transfer can be performed through parallel and serial ports. This transfer is tightly coupled with the processor P1, i.e., Rabbit2000, which we represent using extended notation in UML-RT. Similarly, "SW P1" represents the program executed on processor P1 that handles communication with the controlled object. We can further refine u: User Interface and p: Process Data in this way. Based on the first stage, we combine the previous platform division to perform platform modeling. The modeling of the system's ARC platform is shown in Figure 4. [align=center] Figure 4 UML-RT Modeling of the ARC Platform Figure 4 shows the basic components of the ARC platform and their topology, which can be used for further object modeling. This model provides a framework and allows for easy addition, movement, substitution, and modification of elements within the framework. UML-RT ports effectively represent the interfaces between model elements. The API platform sits between the ARC and ASP platforms and includes the RTOS, embedded web server, device drivers, etc. From a UML-RT perspective, the API platform is the communication channel between the ARC container and the ASP container. Here, the ARC and ASP containers are directly connected through a connector. We represent the API platform as shown in Figure 5. ARCProxy receives call information from the ASP container's port and then sends this information to the appropriate device driver through the RTOS API, allowing the device driver to communicate with the ARC platform. Asynchronous and synchronous communication can be performed in this way. As can be seen from Figure 5, the driver and RTOS are closely related. The modeling of the ASP platform can also be represented in a similar way. [align=center] Figure 5 Implementation of Communication between the ARC Container and the ASP Container[/align] In summary, by combining UML-RT and Platform for the design of embedded controllers, we can see that they can effectively abstract embedded hardware and software and provide good documentation. The system architecture designed based on UML-RT and Platform has good adaptability to new design requirements and design modifications that arise during the design process, and designers can discover and correct errors early. The innovation of this paper is that this paper first introduces UML-RT and Platform in embedded system design theory, and then, in combination with embedded controllers, the author elaborates on the application of UML-RT and Platform ideas in embedded controller system design, and gives detailed steps and specific content of embedded controller system design. References: [1] Kang Jun, Dai Guanzhong. Design of Industrial Ethernet Remote Monitoring System Based on Internet [J]. Control Engineering, 2004.7: 16-20 [2] Wang Yong. Research on the Implementation of Embedded Internet and its Security Issues [D]. Doctoral Dissertation of Zhejiang University, 2003 [3] Li Fengbao. Networked Measurement and Control System Technology [M]. Sichuan University Press, 2004.5 [4] Dong Xiaoguo, Wang Rong. Design and Implementation of Web-based Remote Monitoring System for Reclaimed Water [J]. Microcomputer Information, 2006, 12-1: 86-88