Abstract This paper mainly introduces the adaptive optimization and transformation of the blast furnace blower automatic control system based on modern redundancy and fault tolerance technology, improving the stability and reliability of the blast furnace blower control system and ensuring the safe and stable operation of blast furnace production. 1 Introduction The blast furnace blower (hereinafter referred to as the blower) is a device that provides cold air to the blast furnace for smelting. Its working principle is that the blower is driven by a steam turbine or motor to rotate at high speed, compressing the ambient temperature and pressure air to a certain pressure and temperature, and then supplying it to the blast furnace for molten iron smelting, completing the process of converting steam thermal energy or electrical energy into kinetic energy. The blast furnace blower plays a very important role in the molten iron smelting process and is one of the important factors restricting blast furnace production and smooth operation. At present, Laiwu Steel has 4 blast furnaces, with only 5 blower units. Among them, blowers No. 3 and No. 4 are original Japanese blowers imported 10 years ago, and the equipment and instrumentation and control systems are aging. During normal production, 4 units supply air, with only 1 standby unit. For the wind turbine side, backup power is severely insufficient. If one operating turbine fails and the backup turbine is put into operation, the entire power plant will face a situation where there is no backup power. Therefore, it is essential to strengthen the safe and reliable operation of the existing wind turbine control system. Improving the stability and reliability of the automatic control system is currently the most pressing issue. Redundancy and fault tolerance technology is an emerging technology developed in recent years, possessing advantages such as high reliability, high availability, and no single point of failure, making it very suitable for application in wind turbine retrofitting. In this retrofit, this technology was widely applied, mainly focusing on three aspects: power supply, network structure, and process interlocking parameters. 2. Power Redundancy Optimization The control system does not have very stringent requirements for the quality of AC power, but it has extremely high requirements for uninterrupted power supply. Therefore, the wind turbine control system is equipped with a UPS (Uninterruptible Power Supply). When the mains power fails, it switches to UPS battery power to ensure the control system continues to operate normally for a period of time. The UPS provides a certain guarantee for the stability of the automatic control system, but in actual operation, due to many factors such as the site environment, battery activation, and power grid quality, the UPS still faces many problems during actual switching, leading to various power supply failures. Statistics on UPS failures in recent years show that UPS power outages mainly occur during the switching process between the UPS main circuit, AC bypass, and maintenance bypass. Due to the difference in transient voltage during switching, "circulating current" is generated. Excessive circulating current can cause UPS inverter failure, leading to output power distortion or even momentary power outages. Furthermore, the randomness of the fault during switching makes it difficult to monitor. Therefore, it is necessary to optimize and upgrade the power supply. Redundant configuration of UPS systems would incur significant costs, and both parallel and series UPS connections place high demands on UPS synchronization and the ability to handle step loads. Considering that existing control system power supplies are already redundantly configured, the power redundancy issue can be addressed on the UPS load side. The control approach is: the mains power and the UPS power supply simultaneously power the control system, with the DCS itself handling the power redundancy switching. As shown in Figure 1: [align=center] Figure 1 Simplified diagram of the modified power supply[/align] According to the above design, when the UPS fails, the other mains power supply will continue to supply power to the system normally; when the mains power circuit has a problem, the UPS power supply will still work normally; even if both power supplies fail simultaneously, as long as the UPS battery is working properly, the system can still operate normally for a period of time. This greatly enhances the reliability of the system power supply, reduces power failures, and improves the overall system reliability coefficient. 3 Network Topology Optimization The original communication cable for the No. 1 fan in the thermal power plant was a thin coaxial cable, and the network structure was a bus structure. The connection between the coaxial cable connector and the network interface card often became loose, causing network communication interruptions between the process station and the monitoring station, affecting the operators' monitoring and operation of the unit. The original network only had one host computer. When the host computer crashed, the unit's operation status would not be visible for a short period of time, which is dangerous for high-speed operating equipment. Therefore, the network topology of the unit was modified according to the actual production needs. The network structure diagrams before and after the modification are shown in Figure 2. [align=center]Figure 2 Network Topology Diagram Before and After Modification[/align] A HUB was added to the network, transforming the network structure from a bus network to a star network. Coaxial cables were replaced with twisted-pair cables. Since the control station interface lacked RJ-45 interfaces, a D-Link adapter was used at the process control station to convert the AUI interface to an RJ-45 interface, thus optimizing the entire network. After the optimization, the problem of frequent network communication interruptions was completely eliminated during operation, and the system's security and stability were enhanced. Based on the needs of the site, the network structure of the control system for the No. 5 fan unit was modified as follows: The original topology of the No. 5 fan is shown in Figure 3. The original network could not achieve complete redundancy. When a communication module CI830 failed, the entire slave station containing the failed CI830 would experience a complete communication interruption, which would have serious consequences for production. A case of this module failure occurred at an oxygen generator in Laiwu Steel, causing an oxygen generator shutdown. After evaluation, the network structure was modified, replacing the original CI830 communication module with the CI840 communication module, which has redundant switching capabilities, as shown in Figure 4. This not only simplified the network structure but also made the entire network completely redundant. If any component on the network fails, the entire network will remain operational and will not affect normal production. [align=center]Figure 3 Original Network Topology[/align] 4 Process Interlocking Parameters For high-speed rotating equipment, surge detection is crucial. If the turbine speed is normal but the differential pressure at the fan throat is too small, surge is most likely to occur. In the original backflow judgment, only a differential pressure transmitter was used to judge the differential pressure at the fan throat. Since a single transmitter may experience electrical faults, such as aging, broken wires, or loose terminals, resulting in erroneous signal changes and accidental shutdowns, a strategy of using three differential pressure switches in a three-to-two logic judgment was adopted to eliminate erroneous judgments caused by single-point errors and improve the stability and reliability of the fan. [align=center]Figure 4 Network Topology Diagram After Modification[/align] [align=center]Figure 5 Three-out-of-two Differential Pressure Decision[/align] These differential pressure switch signals are split into two paths via relays: one path enters the DCS to participate in the reverse current judgment in the DCS; the other path enters a small PLC (OMRON sysmac CPM1A). Whether in the DCS or the small PLC, these three differential pressure signals use a three-out-of-two voting method. Once two of these three differential pressure signals reach the reverse current value, a corresponding delay is immediately applied. The outputs of the DCS and PLC simultaneously control the field equipment, causing the unit's anti-surge valve to open or the unit to shut down. This improves the speed and accuracy of reverse current judgment. If the DCS fails, as long as the PLC is normal, after a certain delay, the unit's anti-surge valve can still open or shut down after a reverse current occurs. Thus, the unit can still be protected even if the DCS fails. The logic diagram is shown in Figure 5. For signals that cause process interlock shutdowns, such as low lubricating oil pressure and low power oil pressure, a 3-to-2 logic judgment has been added to the program. Two temperature switches have been added to form a 3-to-2 logic judgment together with the original fan inlet temperature RTD, making high inlet air temperature a condition for safe unit operation, thus increasing the unit's safety performance. 5. Conclusion Statistical data shows that in practical applications, the DCS function is only utilized at less than 30%. Upgrading the existing DCS to achieve advanced control only requires an increase of 10% in cost, while achieving 40% benefits. After optimization using this automatic control system, the power plant reduced the amount of cold air emitted by approximately 100 m³/min, and reduced unplanned shutdown time by approximately 50 hours per year, bringing considerable economic and social benefits of over 2 million yuan. This optimization and upgrade won the first prize for scientific and technological breakthroughs from the Laiwu Steel Automation Department in 2003 and the 2003 Laiwu Steel-level technical appraisal, reaching the advanced level in the province. Its practical application effect is excellent and it is worthy of widespread promotion in the metallurgical industry.