Analysis and Improvement of PLC Communication Interruption in Power Plant Desulphurization
Stationary Hydraulic Dock Leveler Stationary Hydraulic Dock Leveler,Electric Dock Leveler,Hydraulic Leveler,Hydraulic Dock Plates Porema (Suzhou) Logistics Machinery and Equipment Co., LTD , https://www.jsporema.com
First, the desulfurization PLC system configuration situation Wangtan power plant desulfurization PLC control system uses a tree-shaped network, set up two layers of control network: the upper network for the auxiliary workshop centralized monitoring network, the lower network for the desulfurization plant-level control backbone. The whole plant auxiliary control network has 4 operator stations, 1 history station, 1 engineer station and 2 sets of redundant, hot standby servers and redundant switches; the workshop-level control backbone adopts 100M redundant optical fibers. Ethernet, with 3 operator stations, 1 engineer station, 1 history station and redundant switches, is equipped with four PLC control systems: #1FGD, #2FGD, #1-2FG, and #1-4FGD. Equipped with a central processing unit (CPU) 140CPU53414A four sets (8 blocks in total), two sets of hot standby modules 140CHS11000 four sets (8 blocks in total), redundant communication modules 140N0E77101 four sets (total 8 blocks), input and output modules , dedicated connection cables and connectors and real-time operating systems. The PLC system programming software is Concept 2.6, and the monitoring software is Ifix3.5 Unlimited Chinese Development Version. The desulfurization PLC control system communicates with the whole plant auxiliary control network through 1000M redundant optical fiber Ethernet switches. The communication protocol is TCP/IP. Through the communication interface, the monitoring of the desulfurization system is included in the whole plant auxiliary control network, and the unit unit centrally controls the indoor auxiliary. The monitoring station's operating personnel completed the monitoring and management of the two furnace desulfurization systems. The communication network between the operator station and the control station is dual redundant industrial Ethernet, redundant switches, and communication protocol TCP/IP. The communication network between I/O stations uses redundant MODICON RIO networks, ie remote I/O networks. Field system structure diagram 1. 
Second, desulfurization network communication interruption causes analysis There are two servers on the auxiliary control network to directly collect data from all PLCs. In the desulfurization system, there are five host computers that collect data from the PLC. The host computer SCADA software uses IFix3.5. #1FGD, #2FGD, #1-2FGD, and #1-4FGD are Quantum's hot standby systems. The entire desulfurization system uses a German Hirschmann switch for dual network configuration.
The memory data allocation of each station and the data request of the host computer are shown in the following table 1: 
The performance of the host computer communication has a greater relationship with the CPU scan time, data request volume, and the structure of the host computer. From the above table, we can obtain that the program size of other stations except #1-2FGD is relatively large. The scan time is more than 50ms. In addition, from the point of view of the 3:X type of data, the data volume of other stations except #1-2 FGD is more than 50,000 words. These factors cause the PLC to need a scan cycle of about 200ms after the establishment of dual-system hot backup. Because each cycle ensures the synchronization of the main and standby data, these data need to be transmitted from the host to the standby. The on-site inspection #1FGD has an actual scanning cycle of about 196ms after the establishment of dual-system hot backup, which is nearly three times larger than that of the stand-alone, which makes the response to the upper computer very slow. In addition, a total of 7 PCs in the desulfurization system can directly collect data from the PLC, which can also lead to a slower host computer response. When there is a communication timeout, SCADA will show a communication interruption, but at this time, PLC processing for process control is normal. To improve the speed of data response can be analyzed from the above aspects.
Third, the improvement of the feasibility of the program 3.1 to reduce the number of direct access to the host PLC According to the actual operation of the need to retain the appropriate number of host computer, usually not using the station to close its IFix3.5 can improve communication performance. Or use the client/server method to keep two host servers collecting data from the PLC and other operator stations to get data from the server.
3.2 Reasonably configure the host computer data request to reduce the amount of data request In IFix, a request for discrete data can be collected in 2000 points, and for word type data, 125 words can be acquired. When configuring I/O data requests, the data that needs to be collected is placed in the same request and collected to reduce the number of data requests. For example, the 0:X type data of #1FGD, #2FGD, and #1-4FGD can be optimized. One request can be reduced. For 7 PCs, 7 requests can be reduced. However, such changes may require minor modifications to the procedures. In addition, the dual network structure adopted by the field system can allocate the upper computer to collect data from different NOE modules. If the auxiliary control network collects data from a NOE, local control collects data from another NOE, which can improve the SCADA response performance.
3.3 Optimizing the Program When reducing the scanning cycle during dual-system hot backup The large amount of program in the site will lead to a large number of hot spare words required for hot standby. As a result, the scanning cycle for hot standby is greatly increased. It is possible to optimize programs such as reducing the application of non-locating variables, and reducing the application of DFB in dual-system hot standby systems can reduce the number of hot spare words, but this modification workload is large.
3.4 The new Unity Quantum dual-system hot standby CPU module used above can properly improve the current communication performance, but Unity Quantum CPU is used if it needs to greatly improve the communication performance. There are two main reasons: The new CPU's program calculation speed and data transfer speed during hot standby are greatly increased, making the PLC scan cycle very short. The Ethernet communication response request capability under Unity is two to four times higher than the Quantum hot standby under Concept. After converting the #1 FGD program to the Unity program, the scan cycle can be reduced to about 40 ms in the case of dual-system hot backup according to the test result. Without changing the current configuration of the host computer, the theoretical calculation can have 30 PCs connected at the same time to meet the performance requirements. It is more convenient to convert the Concept program to Unity Pro's program. The program structure is similar to Concept and requires only a small amount of checking. UnityPro's interface is slightly different, but it's easy to learn on the basis of Concept. The hardware and wiring of the system need not make any other changes except for replacing the CPU and the CHS module. Therefore, we chose to upgrade the original control system CPU controller to achieve the purpose of reducing the scanning cycle.
IV. CPU Upgrade and Precautions 4.1 CPU Upgrade Based on the above analysis, the final decision is to adopt Option 4, replacing the 140CPU53414 CPU in the original system with 140CPU67160 (requiring 7M memory), and preparing the EthernetMTRJ-MTRJ fiber-optic cable for hot backup. The two CPUs are connected to each other. And the addition of a scalable Unityv2PCMCIA memory card (SRAM) for the newly replaced CPU, model TSXMRPC007M, enables the control system to achieve reliable, redundant, hot-spare. Simply configure the newly replaced CPU and slot number in the Unity programming software. The scanning cycles of the upgraded PLC control systems #1FGD, #2FGD, #1-2FG, and #1-4FGD are only 34ms, 37ms, 19ms, and 40ms, which completely solves the problem of communication interruption during hot standby.
4.2 Notes on CPU upgrade 4.2.1 Process system safety outage During the CPU upgrade process, the operating status of the process system cannot be monitored and controlled. The entire upgrade process will take at least one or two hours, and as many as ten hours or more. It is best to choose when the unit is out of service. If it cannot, then safety measures for the corresponding equipment must be made. Equipment that cannot be stopped should be switched to local operation, such as mixers and lubricant pumps.
4.2.2 CPU model and NOE version match It needs to be emphasized that the model of the CPU must match the version of the NOE. Otherwise, the program cannot be downloaded to the CPU. During the upgrade process, the program can connect to the CPU through the MAC address, but cannot download the program through the Ethernet and USB interfaces, because this upgrade is based on the original concept2.6, and the NOE module is purchased in 2005. Installed and used, obviously the NOE version is inconsistent with the model of the CPU. Before upgrading, you must use the OSLOADER function in the Unity ProXL program to upgrade the NOE module to the matching version in Unity. After the system name, system hardware number, and wrong system version are displayed correctly, the program can be downloaded. .
4.2.3 Configuration of IP Address When Unity ProXL is connected via Ethernet, the four groups of hexadecimal numbers written on the NOE module are first converted into decimal four groups of IP addresses, and then the IP address of this unit is changed to The address of the same network segment can be. Sometimes it is not possible to connect, you can try to configure the last one of the IP on the Ethernet module plus 1, because plus 1 is an IP address with hot standby, the system automatically adds 1.
After the implementation of the plan, a more obvious effect was achieved, and the CPU hot standby operation of the desulfurization system was realized, and no communication interruption occurred. It lays a solid foundation for the stable operation of the unit, comprehensively improves the overall control level of the whole plant auxiliary control system, and lays a solid foundation for the unit's safety, stability and economic operation.