1. Overview Currently, China conducts insulation monitoring of high-voltage equipment according to the "Preventive Testing Regulations for Electrical Equipment," carrying out periodic testing and maintenance. However, preventive testing typically does not consider the equipment's operational condition, requiring mandatory repairs upon expiration, resulting in low effectiveness and sensitivity, and failing to fully meet the safety, economic, and stable operation requirements of the power grid. According to incomplete statistics, 80% of transformer accidents nationwide between 1985 and 1990 occurred even after passing preventive testing. Therefore, there is an urgent need to shift maintenance methods from the traditional time-based approach to a condition-based approach. Condition-based maintenance (CBM) for electrical equipment is receiving increasing attention in power systems, and online insulation monitoring technology, as a prerequisite for implementing condition-based maintenance, has become a research hotspot in the high-voltage field both domestically and internationally in recent years. Practice shows that online monitoring of high-voltage equipment insulation parameters can not only promptly detect latent faults, prevent major insulation accidents, and improve power supply reliability, but also reduce the blindness of equipment outage testing and maintenance. Previous online insulation monitoring systems mostly adopted a centralized processing method, that is, the measured signal was introduced into the system host through a shielded cable, and then the host performed centralized cyclic detection and data processing. Because the primary signal is very small, the measured parameters are greatly affected by interference introduced during analog transmission after sensor coupling, and the validity and stability of the measurement results cannot be guaranteed. With the rapid development of computer and communication technologies, a bus-type online insulation monitoring system can be adopted. In addition to signal extraction, the field monitoring unit also has signal preprocessing, digitization and processing functions, truly realizing the distributed measurement of insulation parameters. This paper will study this. [b]2 Measurement Principle 2.1 Transformer[/b] Theoretical analysis shows that hydrogen is a characteristic gas reflecting transformer faults. Many insulation faults that occur in transformers during operation, such as overheating faults such as multi-point grounding of the core, partial short circuits, and poor contact, as well as discharge or moisture-related faults, all produce hydrogen. By monitoring the dissolved hydrogen in the transformer oil online, latent faults in the transformer can be effectively detected. In addition, multi-point grounding faults of the transformer core can be detected by monitoring the core grounding current online. [b]2.2 Capacitive Equipment[/b] For capacitive equipment such as transformer bushings, current transformers (CTs), capacitive voltage transformers (CVTs), and coupling capacitors, insulation defects can be detected relatively sensitively by measuring dielectric loss (tanδ) and capacitance. The key to realizing online detection of dielectric loss parameters in capacitive equipment is how to accurately obtain and calculate the phase difference between the fundamental waves of the current and voltage signals. Since the traditional zero-crossing comparison method has complex hardware circuitry and poor anti-interference performance, this paper adopts a purely digital method based on Fast Fourier Transform (FFT). Two high-precision current sensors couple the voltage and current signals of the monitored equipment. Then, a digital measurement system performs integer-cycle sampling (A/D) and Fast Fourier Transform on the signals to obtain the fundamental vectors of the two signals and their phase difference, thereby calculating the dielectric loss value. This method does not require complex analog signal processing circuits and can effectively suppress harmonic interference. 2.3 Surge Arresters Monitoring the change in resistive current component of zinc oxide surge arresters (MOAs) during operation is a more effective and sensitive method for determining the degree of valve plate deterioration or moisture absorption. This paper adopts a method similar to that used for capacitive surge arresters for monitoring the resistive current component of MOAs. The fundamental component of the resistive current can be obtained through FFT analysis of the bus voltage signal and the MOA leakage current signal. [b]2.4 Measurement of Surface Leakage Current and Ambient Temperature and Humidity[/b] Ambient temperature and humidity are important external factors affecting insulation parameters. By monitoring conventional climatic parameters such as ambient temperature and humidity, combined with monitoring the surface leakage current of the equipment's porcelain bushings, the degree of contamination of the external insulation of the equipment can be determined, which helps improve the reliability of online monitoring data diagnostic results. [b]3 Bus-type Monitoring System 3.1 Structure[/b] Figure 1 is a schematic diagram of the bus-type online insulation monitoring system. The system consists of three parts: a local monitoring unit (LC), a substation communication controller (SC), and an insulation diagnostic system (IDS). [align=left][b]3.2 Local Monitoring Unit[/b] All local monitoring units consist of a sampling sensor module, a signal conditioning and A/D sampling module, an embedded microcomputer module, and an RS485 communication and power management module, as shown in Figure 2. Signals are directly preprocessed and acquired after coupling with the sensor. Digital signal analysis, processing, and communication functions are all performed on-site by the local monitoring unit. [/align][align=left]3.2.1 Signal Sensor The sensor directly affects the measurement accuracy of insulation parameters. A compensated zero-flux current sensor was designed for capacitive devices to improve the accuracy of small current detection. Permalloy, with high initial permeability and low loss, was selected as the core. Deep negative feedback compensation technology was used to automatically compensate the excitation magnetomotive force of the core, enabling the core to operate in a near-ideal zero-flux state. Simultaneously, the coil was well-shielded with double layers, effectively improving the stability of the sensor's angle difference and ratio difference, and giving the sensor good temperature characteristics. [b]3.2.2 Embedded Microcomputer Module[/b] The embedded microcomputer module is the core of the local monitoring unit. It adopts a new generation of 32-bit microprocessors, which include most of the peripherals of a personal computer and add the features of a microcontroller, enabling it to perform complex mathematical operations and making it very suitable for embedded systems. This system uses a 32-bit CPU as its core and designs an embedded microcomputer system, suitable for on-site measurement, and possessing powerful data processing and port control functions, making signal processing very simple, thus enabling localized on-site monitoring. It is worth noting that traditional insulation monitoring systems introduce the measured signal into the system host via a shielded cable for detection and data processing. Because the primary signal is very small, interference introduced during analog transmission after sensor coupling has a significant impact on the measurement results, which is one of the fundamental reasons for the large measurement dispersion in online monitoring systems. In the local monitoring unit, the sensor output signal does not need to be transmitted over long distances; it is acquired and processed locally, greatly improving the system's anti-interference capability and measurement stability. **3.3 Substation Communication Controller** The substation communication controller controls the LC's operating status and reads its monitoring data via an RS485 bus. It communicates with the insulation diagnostic system through a public communication network. The RS485 bus is a universal field communication bus, a bidirectional, multi-point, digital communication mode with good anti-interference performance and long-distance transmission capabilities, suitable for the communication needs of a substation. Figure 3 shows a schematic diagram of the communication network structure of the insulation monitoring system. The substation communication controller can connect all local measurement units via a single pair of RS485 communication buses (usually twisted-pair cables). Although each LC has independent acquisition, signal processing, and measurement functions, its measurements are controlled by the SC. The SC communicates with the LC according to a specific time and task sequence, starts the LC's data acquisition system, fulfills the requirement for synchronous measurement of insulation parameters, and reads the measurement data and status information of each LC. If one LC fails, the SC can remove it from the network; adding or removing one local measurement unit will not affect the system's operating status. This enables truly distributed local measurement of insulation parameters and bus-based communication for the entire system. [b]3.4 Insulation Online Diagnostic System (IDS)[/b] The NT platform-based insulation online diagnostic system automatically obtains substation monitoring data through the public communication network. Using SQL Server as the database server, it analyzes the changing trends and correlations of various parameters, comprehensively diagnoses the operation and insulation status of each device, and guides equipment operation and maintenance. 4. Field Operation Results Since online monitoring data is affected by various factors such as PT voltage reference, environment, operating mode, and voltage fluctuations, assuming the insulation status of the equipment is good, the data obtained from online monitoring of parameters such as capacitance of capacitive equipment and leakage current of MOA are very stable. However, the data obtained from online monitoring of dielectric loss of capacitive equipment and resistive current of MOA will fluctuate within a certain range. As shown in Figures 4 and 5, the monitoring parameters such as dielectric loss of capacitive equipment and resistive current of surge arresters undergo periodic changes every day. Analysis shows that the data changes are caused by changes in ambient temperature and humidity, which fluctuate periodically every day (see Figure 6). The monitoring parameters showed a strong correlation with ambient temperature and humidity, and the insulation status of the equipment could be analyzed by examining this correlation. Furthermore, for similar high-voltage equipment, since the correlation between the monitored parameters and ambient temperature and humidity is consistent, this principle can be used for further analysis and processing of the data. The influence of ambient humidity on the monitoring data is mainly reflected through the leakage current on the porcelain bushing surface (see Figure 7). For example, since the leakage current on the outer surface of the surge arrester porcelain bushing enters the detection circuit, when the bushing surface is dirty and the environment is humid or rainy, the current flowing through the bushing surface will cause the resistive current monitoring result to be significantly higher; similarly, the current on the bushing surface will also have a certain impact on the dielectric loss monitoring result of some capacitive equipment. From the above analysis, it can be seen that the online data measured by the monitoring system is accurate and reliable. Although the data has some fluctuations, these have their own inherent causes and are not caused by the monitoring system itself. This also reflects the characteristics of online monitoring data, indicating the necessity of establishing an online expert diagnostic system to comprehensively analyze the online monitoring data in order to determine the insulation status of the equipment in operation. [align=left][b]5 Conclusion[/b] An online insulation monitoring system based on an embedded microcomputer system and an RS485 bus network structure was proposed. Signal extraction, acquisition, processing, and analysis are all independently completed by the field microcomputer unit, realizing localized insulation monitoring. The results show that, by achieving local measurement of insulation parameters, long-distance transmission of small signals is avoided, the problem of data dispersion in online monitoring is solved, and the stability and anti-interference capability of the system measurement are significantly improved.