Abstract: This paper mainly introduces a control system based on the popular fieldbus DeviceNet and CompoBus/D, and elaborates on the application of a three-level computer communication network architecture and industrial control network technology. Based on this network, the author uses both fuzzy control and PID control methods to control the same object (an electric oven). When constructing the control system, the author considers both performance and cost. The results show that the control effect achieved using the DeviceNet fieldbus is very ideal, and the fuzzy control method is superior to the PID control method for the electric oven. Keywords: DeviceNet; CompoBus/D; PLC; fuzzy; PID; temperature; water level Abstract: This paper introduces a control system based on the fieldbus of DeviceNet and CompoBus/D, which are currently popular fieldbuses. It describes a three-level communication network architecture composed of two kinds of fieldbuses and the application of industrial control networks. During the design of the control system, the author takes into account both the performance and cost of the system. The author controls the same object with two kinds of methods, namely fuzzy control and PID control. The results show that the control effect based on DeviceNet fieldbus is fine, and the fuzzy control method is superior to the PID control method in controlling an electronic oven. Key words: DeviceNet; CompoBus/D; PLC; fuzzy; PID; temperature; water level 1 Introduction Rockwell has created a unique three-layer industrial control system network structure composed of DeviceNet (device layer), ControlNet (control layer), and EtherNet (information layer). The CompoBus/D network is a standard fieldbus from Omron, belonging to the equipment production line control level network. This bus is also a fieldbus based on DeviceNet. This paper proposes to improve the cost-effectiveness and reduce costs while focusing on performance. It proposes to use the openness of DeviceNet fieldbus and adopt products from different companies to build a control system, thereby improving the cost-effectiveness and reducing costs. This paper mainly discusses: (1) Industrial communication between ROCKWELL PLC and Omron PLC and frequency converter through DeviceNet fieldbus. (2) Remote control through industrial Ethernet. (3) Application of fuzzy control theory and PID control 2 System Hardware Structure Figure 1 is the system structure diagram with configurator. The system can support up to 64 nodes and achieve remote control by communicating between the master unit and the slave unit. (1) The top layer is industrial Ethernet, which consists of a computer, hub, and ROCKWELL SLC5/05 PLC. The computer is mainly used for remote monitoring, and the PLC is mainly used as the control master station to transmit the computer's control information to the lowest device layer and transmit the device information to the desktop computer; it can also perform control functions itself. The SLC5/05 comes with an Ethernet interface. ROCKWELL also provides Ethernet modules. (2) Composition of the master station PLC: The ROCKWELL SLC5/05 PLC is used as the master station. The modules it comes with are: CPU: 1747-L551, analog input module 1746-NI8, analog output module 1746-NO4I, digital input module 1746-IB16, digital output module 1746-OB16, digital output module 1746-OW16, DeviceNet scanning module 1747-SDN. The DeviceNet scanning module 1747-SDN is responsible for the communication of the underlying DeviceNet network. (3) The device layer consists of two control objects. One is the Omron PLC, which is mainly used to control the electric oven. The Omron PLC consists of the following modules: CPU module OMRON CQM1H-CPU51, Analog input/output module OMRON CQM1H-MAB42, Digital input/output module CQM1H-ID212, Digital output module OMRON CQM1-OC222, and CompoBus/D scanning module OMRON CQM1-DRT21. The electric oven uses a 4-20mA current sensor as the temperature measurement signal, connected to the analog input terminal of the Omron PLC, and uses a 4-20mA current signal as the output signal. Secondly, there is the Omron frequency converter CompoBus/D communication card. This card is an optional device for the frequency converter. The liquid level sensor is a 4-20mA current sensor, connected to the analog input terminal of the Omron PLC. The entire system structure is shown in the following figure: [align=center] Figure 1 Network System Structure[/align] 3 Implementation of Data Information Exchange After completing the hardware wiring, three more parts need to be completed here. First, configure the master ROCKWELL PLC using RSLogix software. Next, configure the Omron PLC using CX-PROGRAM software. Then, configure the DeviceNet network using the DeviceNet configurator. The first step is to set the address and communication baud rate for each DeviceNet slave. The communication baud rate for the entire network must be consistent. Then install the RSNetworkx software and the DeviceNet configurator hardware. I chose the ROCKWELL 1770 KFD. Because it's a non-ROCKWELL product, when scanning for OMRON PLCs and inverters, it doesn't recognize these products at all, displaying two question marks when scanning OMRON products. Installing the OMRON product's EDS file allows DeviceNet to recognize the product. (1) EDS file not installed (Figure 2): [align=center] Figure 2 Configuration screen without EDS file installed[/align] (2) After successful DeviceNet configuration, the following is seen (Figure 3): [align=center] Figure 3 Screen after successful configuration[/align] RSLogix 500 is used to program the ROCKWELL PLC, and CX-Programmer is used to program the OMRON PLC. Data transmission between the upper and lower computers can be achieved with just a few simple instructions. In the computer, KingSCADA is used to create the configuration screen, and variables are bound through RSLinx's OPC service to achieve remote control. 4 Control Implementation 4.1 Control of Electric Oven 4.1.1 Control Method of Electric Oven Since the electric oven is a system with large inertia, pure time delay, and nonlinearity, conventional control based on precise mathematical models is difficult to guarantee the heating curve requirements, such as PID control. Therefore, fuzzy control is considered. Let Et be the internal temperature error of the oven, E't be the rate of change of error Et with time, and Ct be the control heating effect. Let PB, PM, PS, P0, N0, NS, NM, and NB represent the values ​​of error Et as positive large, positive medium, positive small, slightly larger than zero, slightly smaller than zero, negative small, negative medium, and negative large, respectively. Let PB, PM, PS, 0, NS, NM, and NB represent the values ​​of E′t and Ct as positive large, positive medium, positive small, zero, negative small, negative medium, and negative large, respectively. Based on experience in controlling the temperature of an electric oven, the following control table (Table 1) can be derived: Table 1 Fuzzy Control Rule Table for Electric Oven Define the fuzzy subsets of Et, E't, and Ct as: {Et}={NB, NM, NS, N0, P0, PS, PM, PB} {E't}={NB, NM, NS, 0, PS, PM, PB} {Ct}={NB, NM, NS, 0, PS, PM, PB} Define its universe of discourse as: {Et}={-6, -5, -4, -3, -2, -1, -0, +0, +1, +2, +3, +4, +5, +6} {E't} = {-6, -5, -4, -3, -2, -1, 0, +1, +2, +3, +4, +5, +6} {Ct} = {-7, -6, -5, -4, -3, -2, -1, 0, +1, +2, +3, +4, +5, +6, +7} The membership degree of each fuzzy variable to its corresponding universe of discourse is described using a normal distribution. Based on the fuzzy control calculation rules, the fuzzy relation matrix can be calculated first: Rt = R1∪R2∪…∪Rk (k = i, j) Where: RL = (Et(i) × E't(j))·Ct(i, j) (i = 1～8, j = 1～7, L = 1～i× j) Based on the above fuzzy control rules, the Ct control matrix is ​​calculated. According to the principle of taking the largest membership function, the corresponding fuzzy control quantity can be obtained. The following is the fuzzy control table (Table 2): The selection of the actual control quantity can be based on the actual measured Et and E't. After fuzzy processing, Ct is obtained by looking up the table and then converted into the actual precise control quantity before output. All fuzzy processing can be completed in the computer through the configuration software and VB program interface. However, considering that the main function of the host computer is monitoring, and in order to reduce communication volume, shorten communication delay and reduce the load of the host computer, this system is implemented in the PLC through statements, which is slightly cumbersome. The PLC scanning speed is very fast and there will be no delay. Table 2 Fuzzy Control Table 4.1.2 Control Effect The effects of fuzzy control and PID control on the electric oven are shown in Figure 4 and Figure 5 respectively. It can be seen from the comparison that the fuzzy control effect of the electric oven temperature is better than that of PID control. [align=center] Figure 4 Dynamic response curve of electric oven temperature fuzzy control Figure 5 Dynamic response curve of electric oven temperature PID control[/align] 4.2 Water Level Control PID control is used to control the water tank level. The detailed control principle is shown in the diagram below: To improve operational accuracy and reliability, all PID calculations are performed on the main unit, with the Omron PLC acting merely as an access device. It transmits signals to the host computer via the DeviceNet bus, and the inverter receives control commands from the host computer via the DeviceNet bus. The water tank's pressure sensor provides a 4-20mA water level signal, and the control output is fed back to the Omron PLC's analog input. Note that PID instructions are only for integers; floating-point numbers are not allowed. Therefore, if a floating-point number is input, a conversion from floating-point to integer will occur. The diagram below shows the ladder diagram for inputting PID commands. The control block length is fixed at 23 bytes. The process variable is the address of the cell storing the process input value; this address can be the location of the analog input word storing the input A/D converter value. The control variable is the address of the cell storing the PID instruction output. It is usually an integer value. Detailed parameter settings for PID instructions can be found in the Rockwell documentation. RSTune loop tuning software was used during PID parameter tuning. It allows for convenient, rapid, and accurate tuning of the PID control loop without additional programming. When tuning parameters using RSTune software, it needs to be used in conjunction with ladder logic programs in RSLinx and RSLogix. By creating corresponding topics in RSLinx and using relevant PID instructions in the ladder logic (where the initial values ​​of the PID parameters are set), the real-time values ​​are visually reflected in the interface provided by RSTune during program execution, facilitating the analysis of the control system's performance. Simultaneously, RSTune automatically tunes the PID parameters based on the collected data. The tuned parameters are: KC=1.1, TI=0.1, TD=0.03. 5. Conclusion The control system has been successfully tested and is now running. The host computer uses configuration software to create a human-machine interface and saves control information from the network for future management. By using products from two companies—ROCKWELL's high-end products, which offer superior performance but are expensive, and OMRON's low-end products, which offer better value for money and are cheaper—ROCKWELL's products are used as the core control master station, while OMRON's products serve as non-core slave stations. This approach balances performance and cost. Furthermore, we can see that fuzzy control is superior to PID control for controlling the electric oven. The author's innovation lies in leveraging the extensibility of the DeviceNet network, employing products from different companies in a balanced way, thus improving the overall system's cost-effectiveness and providing a new approach to cost reduction. References 1. Ling Kong, Jiang Shiqin. Microcomputer Information, 2003, Vol. 19, No. 9, pp. 27-28. 2. Allen-Bradley. SLC 500 DeviceNet Scanner Module 1747-SDN, August 2000. 3. Allen-Bradley. Ethernet SLC 500 Processors, August 5, 1997. 4. OMRON C200HW-DRT21 CQM1-DRT21 DRT1 Series DeviceNet Slaves OPERATION MANUAL, 1998. 5. Liu Shuguang, Wei Junmin, Zhu Zhichao. Fuzzy Control Technology, China Textile Press, June 2001. 6. Allen-Bradley. RSTune User Manual [M].