Abstract: This paper introduces the design idea of ​​the frequency converter based on industrial Ethernet communication. Based on the analysis of the communication mechanism of the network frequency converter, an industrial Ethernet communication control strategy based on data communication priority classification is established. The system uses the PID algorithm of the host computer to calculate the output frequency of the frequency converter according to the water level signal transmitted by the network, thereby controlling the water level of the dual-tank water. Experimental results show that the frequency converter can be effectively controlled to control the water level of the dual-tank water through this industrial Ethernet communication. Keywords: network control; industrial Ethernet; frequency converter; DSP 1 Introduction The development of frequency converters is very rapid and they have been widely used in various sectors of the national economy. Most of the existing frequency converters use RS232/RS485 as the communication method. These communication methods have disadvantages such as low communication reliability, slow communication speed, small communication data capacity, short communication distance, difficulty in forming a network system, and low real-time performance, which cannot meet the actual control requirements [1]. Commercial Ethernet has the advantages of high data transmission rate, low cost, and easy networking, but its deterministic transmission problem and real-time problem make it unsuitable for industrial sites [2]. To address this issue, this paper proposes an industrial Ethernet communication mechanism based on data communication priority classification and applies it to a self-developed frequency converter. This industrial Ethernet allows for remote monitoring of the frequency converter, enabling PID water level control of a dual-tank system. 2. Design of the Industrial Ethernet Frequency Converter 2.1 Network Communication Principle of the Frequency Converter Traditional Ethernet uses a bus topology and the IEEE 802.3 CSMA/CD protocol. In industrial control applications with extremely high real-time requirements, the main problem is the difficulty in guaranteeing the maximum latency requirement for data transmission. Figure 1 shows the industrial Ethernet network structure. To address this problem, this paper constructs a topological industrial Ethernet network structure, placing the frequency converter within this network structure for communication control, as shown in Figure 1. This network structure draws on some advantages of fieldbus to improve the real-time performance of Ethernet control data transmission. The system uses a network hub controller as the master node, and the frequency converter and other nodes communicate under its control. The master-slave communication method effectively avoids conflicts. Meanwhile, different data should be classified according to their functions and reliability requirements, and different types of data should be processed separately to ensure both the reliability of control information transmission and the throughput of the entire network. This paper divides the data in the control network into three categories: (1) Real-time and reliable data. This type of data is given a priority of P=1, such as control signals. (2) Real-time data, but not requiring reliability, such as multimedia data. This type of data is given a priority of P=2. If necessary, some data frames can be discarded. (3) Data that does not require real-time or reliability, such as uploaded PLC programs. This type of data is given a priority of P=3. The specific communication strategies for these data packets are as follows: (1) For data packets with P=1, regardless of how many unsent data are in the queue, it must be ensured that the data is sent completely, even if the period exceeds the sending period T. (2) For data packets with P=2, the sending can only be started when the data with P=1 is completely sent. (3) For data packets with P=3, the sending can only be started when the data with P=2 is completely sent. (4) Each sending is coordinated by the network hub controller, which sends an instruction to the designated frequency converter to allow it to send. The designated device receives the instruction and starts sending. Due to the broadcast characteristics of the shared network, the controller is notified at the same time that the host has given up the right to send. (5) In order to ensure the fairness of each station, except for the priority of P=1, the data transmission of other priorities always starts from the next station after the last data sent in the previous cycle. (6) For data with priority P=1, if there is no data to send, a 64-byte blank frame should also be sent to notify the controller that it has given up the right to send. The hardware of network communication is mainly composed of the main control chip TMS320F2812 and the network chip RTL8019AS. The interrupt of RTL8019AS is connected to INT1 of DSP, so the program to implement the network communication function is mainly completed in the INT1 interrupt program. 2.2 Implementation of frequency conversion function The system uses the DSP chip TMS320F2182 of Texas Instruments (TI) as the main control chip, and together with the intelligent power module, it realizes AC-DC-AC frequency conversion. The main frequency conversion circuit consists of rectifier circuit, filter circuit, intelligent power module (IPM) and sampling circuit. The rectifier circuit is composed of a three-phase bridge circuit with diodes to convert the power frequency power supply into DC voltage, and at the same time, DC reactance and capacitor are used to absorb the pulsating voltage. The DC voltage after rectification and filtering is directly supplied to the IPM module for inversion [3]. The IPM module contains gate drive control, fault detection and multiple protection circuits. The built-in current sensor monitors the main circuit of the IGBT, and the internal fault protection circuit detects faults such as overcurrent, short circuit, overheating and undervoltage of the control power supply, which is used to prevent damage to the power module caused by mutual interference or overload of the system. The DSP uses the SPWM method to generate the drive signal, which drives the IGBT after high-speed optocoupler isolation. The system generates the PWM signal by setting the comparison unit and dead time unit in the event manager EVA module of TMS320F2182 [4]. The specific implementation is as follows: (1) Set the comparison control register and comparison mode control register of the comparison unit to determine the comparison output enable, the overload condition of the comparison value and mode register, the use of space vector PWM, and the output mode of PWM pin; (2) Assign an initial value to the counter T1CNT of Timer 1, and correctly set the period register T1PR and the control register T1CON of Timer 1; (3) Correctly set the dead time control register DBTCONA. When the timer count value of the comparison unit is equal to the comparison value in the comparison register, a jump will be generated on the two PWM pins of the comparison unit, and a comparison interrupt request will be issued after 1 CUP clock. When Timer 1 is set to continuous increment mode, an asymmetric PWM waveform is generated. It can be seen that changing the comparison value can change the duty cycle of the PWM wave, and changing the period value can change the frequency of the PWM wave. At the same time, the sampling circuit uses the 8-channel 12-bit A/D converter of F2812 to perform A/D sampling of the external input analog signal[5]. 3. Test on Water Level Control of Dual-Tank Water Tank by Network Inverter 3.1 System Analysis This system monitors the inverter parameters and water tank level through Ethernet host computer software. When the network inverter is running, it sends the frequency and water level signals to the host computer through the Ethernet network established by the system. The system sets the predetermined water level H through the host computer software, using the inverter output frequency f and the water level signal h of the first water tank as input and output parameters, and models the system with the dual-tank water tank as the control object. The system uses the PID algorithm to calculate the model of the dual-tank water tank, obtains the system output signal, i.e., the inverter output frequency, and sends it to the network inverter through the Ethernet network to control the water tank level. The control structure diagram of the test system is shown in Figure 2: Figure 2 Structure diagram of the dual-tank water level control system For the first-stage water tank in the system, we can derive the following relationship: 3.2 PID Water Level Control Experiment The system sets the control water level of the water tank to 25cm through the host computer software, and sets the parameters KP=2.0, KI=0.3, KD=0.5. The host computer uses a PID algorithm to derive the inverter's output frequency value and transmits it to the network inverter via Ethernet. Simultaneously, the network inverter samples the water level signal and sends the sampling results and frequency signal together to the host computer in real time via industrial Ethernet. The water level curve can be obtained from this real-time data. Figure 3 shows the water level curve of the PID experiment (unit: cm/s). As can be seen from the curve, the industrial Ethernet inverter can monitor data and achieve water level control through industrial Ethernet, with good results. The ripple in the water level curve is caused by the interference of eddies in the water tank on the water level sensor; the actual water level remains basically constant. 4 Conclusion Based on the experimental curves, the water level and frequency curves of the system's PID experiment when KP=2.0, KI=0.3, and KD=0.5 can be obtained. The water level curve of this experiment shows that the system has first-order characteristics; after reaching stability, the fluctuations in water level and frequency are relatively small, and the system's control characteristics are relatively ideal. During the experiment, the industrial Ethernet communication function enabled the inverter to effectively realize parameter and real-time data, laying the foundation for remote fault diagnosis. Similarly, network frequency converters can also be applied to remote network control of other advanced control laws, and have broad application prospects in the field of process control. The innovative points of this paper are: 1. The system design implements a frequency converter capable of high-capacity data communication based on industrial Ethernet communication. Through this data classification-based industrial Ethernet communication, the network frequency converter can receive network commands and upload relevant parameters. 2. The system models a dual-tank system, receives the water level signal transmitted by the network frequency converter through host computer software, calculates the output signal of the control system using a PID control algorithm, and controls the network frequency converter through an industrial Ethernet network, thereby realizing the water level control of the dual-tank system. The direct economic benefits generated by the research results of this paper reach 500,000 yuan, and it will have a more profound application significance and value in advanced process control systems. All data in this paper comes from research and development experimental results. The research method uses a dual-tank water level control system as an experimental platform to test the system's performance parameters and obtain experimental data. References [1] Meng Lingdong, Zeng Zhixin, Li Yong, et al. Design of braking system based on DSP AC servo motor control [J]. Microcomputer Information, 2006, (14): 153-155. [2] Henning C. Why industrial Ethernet [J]. CONTROL ENGINEERING, 53 (5): 14-14. [3] Zeng Min, Yang Jiuming, Zhang Quanhong, et al. Research on DC frequency conversion control system based on DSP [J]. Power Electronics Technology, 2005, 39 (6): 112-113. [4] Tang Yongqi. Frequency conversion speed control system based on DSP TMS320F2812 [J]. Instrumentation Technology and Sensors, 2006, (3): 46-48. [5] Zhou Xuesong, He Yanmin, Ma Youjie, et al. Application of TMS320F281 in multi-channel synchronous AC sampling of power system [J]. Automation and Instrumentation, 2005, (4): 39-42.