Condition monitoring and fault analysis of power systems - News - Global IC Trade Starts Here Free

SMD aluminum electrolytic capacitor

inductance

First, the significance of power system condition monitoring First, the establishment of historical records of equipment operation, so that the data and data appearing in the situation of equipment operation are accumulated for later use.
Second, determine whether the operating state of the device is normal or not, and judge the nature and extent of the device failure. The main basis for the judgment is the historical archives established before, including the level of the running state of the equipment, and the features displayed in the process of such failures.
Third, in order to be able to provide the necessary basis for the maintenance work during the implementation of the state maintenance, it is necessary to evaluate the operational status of the equipment, analyze the status, and classify the assessment to form a certain evaluation standard. The main content of the assessment of state detection includes: evaluating the operating state of the equipment, estimating the abnormal state of the cup, and predicting future changes in the fault state of the equipment. The inclusion of these contents in the evaluation system is mainly to provide certain conditions for evaluation, so as to continuously improve and improve the evaluation and monitoring.
In summary, the operational data of the equipment can be continuously accumulated, improved and improved in the state monitoring process, breaking through the constraints of the past management system and improving the management system. Therefore, the author believes that in the modern power system equipment management, the state monitoring system has a non-negligible role.
Second, the study of key technologies for condition monitoring First, the so-called online monitoring system for power equipment in signal acquisition, its function is to continuously check and judge the state of the equipment, and predict the development trend of equipment status; system operation The time is the life of the device, which means that it must be monitored as long as it is still in use.
The acquisition of the state information of the diagnostic object is the primary task of the device running state quantity reflecting the operation of the device. The content of the information includes the voltage, current, frequency, and partial discharge of the power device, as well as the density of the magnetic field line and the normal signal. And fault signal. Typically, the method of acquiring the signal will vary with the characteristics of the signal characterizing the state of the device. There are several methods for signal sampling:
1. The length of the sample of each acquired signal is the length required to process a sufficient amount of data. We refer to this sample as a one-time sampling.
2. The sampling time is well specified in advance, and the sampling frequency is a set period, which is simply a timing sampling.
3, automatic sampling, sampling time is random, sampling is a means of sudden signal mutation.
4. Special sampling, sampling method varies according to the requirements of the diagnosed fault, such as speed tracking sampling, peak sampling and so on.
Second, the data transmission signal processing system is usually far away from the monitoring equipment. Therefore, the data is susceptible to interference, loss and phase shift during transmission (subject to environmental factors), and the data must be firstly converted and pre- Processing and compression packaging, and then transmitted to the processing control center via the communication path. Communication equipment has been widely used in the power field, and optical fiber transmission digital signals can better suppress interference and ensure signal quality.
Third, after receiving the state quantity data packet transmitted by the communication line, the data processing industrial control data processing center uses various mathematical methods to unpack the data. For example, spectrum analysis converts time-domain continuous time signals into different frequency-frequency signals for analysis; in the time domain, correlation analysis between two signals is used to search for processed data of another signal; wavelet analysis; neural network; artificial intelligent. The application of digital information technology and intelligent technology to the data processing of power equipment monitoring systems makes online monitoring of power equipment more real-time and accurate.
Third, the fault diagnosis recommendations First, the use of multi-sensor technology and information fusion processing technology to diagnose a fault phenomenon with a different fault. Multi-sensing technology uses multiple sensors to observe the same object from multiple sides and multiple angles, that is, multiple fault characterizations for the same fault, multi-level and multi-domain (time domain, spatial domain, frequency domain) to collect different feature quantities, select faults Reflects the amount of state information with high sensitivity, which makes it more comprehensive to analyze and diagnose faults.
Information fusion technology is a data processing process that analyzes and synthesizes data from multiple sensors according to certain criteria. There is an inherent relationship between different reflections of different feature spaces due to the same equipment failure. The use of fusion technology to "seek the same dissimilarity" can improve the accuracy of power equipment status detection and fault diagnosis. However, the basic theory of information fusion is still not perfect, and the diagnostic method remains to be studied.
Second, based on the feature space vector fault diagnosis method, the fault feature quantity can be corrected in real time by learning the fault error. This diagnostic method has certain adaptive capabilities and is suitable for fault diagnosis of complex objects with uncertainties and slow time-varying. The essence is to use each fault symptom vector as a new a priori symptom vector in the original a priori symptom vector, and correct the fault feature vector according to the adaptive algorithm. When the fault a priori symptom vector is uncertain, the first fault needs to be manually determined.
Third, for the inherent characteristics of power equipment and the uncertainty caused by insufficient information on the status of online monitoring, consider the principle of maximum membership in fuzzy theory to diagnose the cause of the fault, determine the type of fault, and combine the state signal with the fuzzy mathematical method. Analyze the randomness and ambiguity of the fault.
In addition to the above methods, it is also possible to diagnose faults by combining artificial intelligence, expert systems, neural networks, and the like.
Conclusion In the development of the power system in the last decade, the state monitoring technology and fault diagnosis technology of the equipment as a new technology continue to advance rapidly. Whether it is from the perspective of development prospects or from the perspective of application, it shows a good momentum of development. Although the development of these two technologies in China has continued for quite some time, and various detection devices have been put into production and use, however, the use of condition monitoring and fault diagnosis techniques has not been popularized, and There are some problems that cannot be ignored in both the technical understanding and the use process. We should continue to vigorously explore this technology to improve the stability and efficiency of the power system.