Distributed Control System

The Distributed Control System (DCS) is an automation control system based on microprocessors, with decentralized control functions and centralized display and operation. It is also known as the distributed control system. Its core architecture consists of the process control level and the process monitoring level. The hardware is composed of field control stations, operator stations, and engineer stations. Through a hierarchical structure, it achieves decentralized and centralized management of risks.
The development of the DCS system has gone through three key stages. Before the mid-1970s, it laid the foundation for the decentralization improvement of centralized computer control systems and the technical basis; from the mid-1970s to the mid-1980s, the first generation of DCS (TDC-2000, CENTUM, etc.) was introduced, which achieved decentralized control units and basic monitoring functions, but had limitations such as communication compatibility; the second generation was upgraded to a local area network architecture, with product standardization and modularization development, significantly improving reliability and flexibility. From the mid-1980s to the 1990s and beyond, the third generation achieved breakthroughs in open communication and intelligent functions, and the fourth generation was centered on “information and integration”, building a four-layer system; the latest trend integrates AI and digital twins, such as the iNICS system, which realizes fault warning and intelligent operation, and the system is developing towards an open application ecosystem. [6-7]
This system is widely used in fields such as power, metallurgy, and petrochemicals, with a penetration rate of over 95%. The market size in 2024 reached 14.8 billion yuan, the domesticization rate exceeded 45%, and local enterprises accounted for 40.4% in the petrochemical sector. Typical cases include the Dragon Scale platform (safety level) and Dragon Fin platform (non-safety level) used in the Linglong No.1 nuclear reactor, which are responsible for reactor safety control and operation management respectively. On April 10, 2024, the first DCS cabinet was installed and started the installation and commissioning work. Maintenance measures include regular maintenance, grounding resistance testing, and power redundancy experiments, ensuring that the system utilization rate is close to 100% and the average downtime exceeds 100,000 hours.
System Introduction
Announcement Editor
System Name
The term “distributed control system” was derived by translating the product name of a foreign company. Due to the numerous manufacturers and varying system designs, as well as the diverse functions and features, the naming of these products also varies greatly. In China, there are different names when translating, and the most common ones include distributed control system (distributed control system, DCS), total distributed control system (total distributed control system, TDCS), and distributed computer control system (distributed computer control system, DCCS). [6]
The differences in names merely reflect variations in naming intentions and translations. Their fundamental nature of the system remains the same, and their intrinsic meanings are consistent. In the Chinese power industry, it is commonly referred to as a distributed control system. [6]
System Explanation
DCS usually adopts a hierarchical structure with multiple levels. Each level consists of several subsystems, and each subsystem achieves several specific limited goals, forming a pyramid structure. Reliability is the lifeblood of DCS development. To ensure the high reliability of DCS, there are mainly three measures: first, widely apply highly reliable hardware equipment and production processes; second, widely adopt redundancy technology; third, widely implement fault-tolerant technology, self-diagnosis of faults, and automatic processing technology in software design. Currently, the MTBF of most distributed control systems can reach tens of thousands or even hundreds of thousands of hours.
The distributed control system is a distribution of multiple physical resources and logical resources (multiple computers or processing units, multiple data sources, multiple instruction sources and programs), using a certain interconnection network or communication network to interconnect resources, having a high degree of local resource autonomy ability, mutual cooperation ability among resources, and overall coordination and control ability of resources. It can achieve dynamic management and allocation of distributed resources, parallel operation of distributed programs, and computer network control systems with decentralized functions. The meaning of the distributed control system is mainly reflected in “dispersion”, and the meaning of “dispersion” includes several aspects. [6]
Dispersed configuration
The locations of each controlled device are scattered, and the corresponding system control equipment is also configured in a dispersed manner. Multiple distributed control units based on microprocessors are respectively responsible for different control tasks. [6]
Function dispersion
The functions possessed by the distributed control system are not concentrated in a single central control unit. Instead, they are distributed among various decentralized control units. Moreover, the functions such as data acquisition, process control, operation display, monitoring operations, and self-tuning in the control system are also dispersed and relatively independent. [6]
Display dispersion
The display function of the decentralized control system can not only be centralized on the central operation station, but also be distributed to local operation stations. The central operation station has the ability to display all the information of any decentralized process point in the entire system, and can be displayed separately on different display terminals. The local operation station can not only display on-site information through the on-site control unit at any time, but also the third and fourth generation decentralized control systems can call information from other local operation stations or the central operation station at any local operation station for decentralized display. [6]
Database decentralization
Modern distributed control systems mostly employ distributed database systems. The field control units and control stations have local databases, which are shared by the entire system. [6]
Communication decentralization
The distributed control system adopts the communication technology of local area networks. Each process unit in the network has equal communication control rights, enabling decentralized communication. [6]
Power supply decentralization
The decentralized control system provides independent power supply devices for different control units, enabling the power supply of the system to be decentralized and enhancing the reliability of the system. [6]
Load distribution
In the distributed control system, the overall tasks are reasonably distributed to each control unit. One control unit only undertakes the control tasks of several local control loops or subsystems. The workload of the entire system is distributed, and the load of each control unit is basically uniform. [6]
Risk dispersion
The realization of “dispersion” means the dispersion of risks throughout the entire system. [6]
Background and reasons for occurrence
Broadcasting Editor
Demand-driven
The complex process control requirements of modern large-scale industrial production provide the core driving force for its development; [6]
Technical Foundation
Based on the summary of the advantages of conventional analog instrument control and early computer control, it has been developed through the comprehensive application of modern scientific and technological achievements. [6]
Core technology breakthrough
In the early 1970s, a major breakthrough occurred in microelectronics technology. The development of large-scale integrated circuits and the emergence of microcomputers and microprocessors provided semiconductor chips and computer systems with small size, strong functionality, high reliability, and low cost, laying a solid material foundation. [6]
Supporting technologies
The further development of CRT display technology and digital communication technology has provided more complete research conditions; [6]
Theory and Design Guidance
Guided by theories such as cybernetics, information theory, and systems engineering, the research and development is carried out with the goal of comprehensive automation, following the design principles of decomposition autonomy and comprehensive coordination. The emergence of distributed control systems is the result of the “4C” technology, a product of the integration and comprehensive development of multiple disciplines, and also an important achievement of the third industrial revolution (computer development). [6]
Development History
Announcement Editor
The distributed control system has evolved through multiple stages since its inception in the mid-1970s. The core features of each stage are as follows:
Mid-1970s (First Generation)
Representative products included the TDC-2000 from Honeywell in the US, the Spectrum system from Foxboro, the Network-90 system from Bailey, the Teleperm-M system from Siemens in Germany, the 900/TX system from Nippon Chuo Corporation in Japan, and the CENTUM system from Yokogawa. The system consisted of five main components: data acquisition devices, field control stations, CRT operation stations, high-speed data communication paths, and monitoring computers. This stage’s products had the advantages of centralized computer control systems, with effective decentralization of control units, and CRT operation stations featuring rich graphics and full system alarm, diagnosis, and management functions. However, the monitoring computer mainly handled more management and information processing functions, using 8-bit or 16-bit microprocessors, with communication being a primary industrial control local network. The dedicated communication protocol limited system compatibility, and some systems lacked sequential control functions. The technology had limitations.
Early 1980s and mid-1980s (Second Generation)
Representative products included Honeywell’s TDC-3000, Taylor’s MOD300, Hitachi’s HIACS-3000, Westinghouse’s WDPF, and Yokogawa’s YEWT-TORIA, among others. Some were new designs, while others were upgrades from the first generation. The system consisted of local area networks, multi-functional field control stations, enhanced operation stations, main computers, system management stations, and inter-network connectors. The core feature was that the local area network served as the system’s backbone, with each unit as a network node; network protocols gradually unified with MAP standards or were compatible with MAP, data communication capabilities improved and moved towards standardization; product design became standardized, modular, and structured; control functions were complete, user interfaces were friendly; management functions were decentralized, significantly improving system reliability, adaptability, and expansion flexibility.