基于多信息融合的交流接触器状态评估方法研究

发布时间：2018-10-23 17:26

【摘要】：交流接触器作为开关电器,它的开断比较频繁,同时它还可以作为一种控制电器,它控制的对象可以是电动机,也可以是电热器、电焊机和电力机组等电力负载,在低压配电系统中运用的非常广泛,是一种非常重要的低压电器,因此其安全稳定的运行对电网的安全稳定运行也是十分必要的,为了保证其安全稳定的运行,就是能够实时的监测其状态并能够做出及时有效的反应。近年来,随着在线监测技术的日趋成熟,基于大数据对电器状态进行评估的方法也越来越普遍。因此,本文以CJX2-80为例,从试验平台设计、特征参数提取及去噪处理、状态评估等方面进行课题的研究。首先,本课题设计并搭建了交流接触器全寿命试验平台,试验平台包括线圈驱动电路、信号采集电路、信号调理电路等,测试系统软件采用的是LabVIEW图形化编程语言进行模块化设计,其是美国国家仪器公司提供的虚拟仪器开发平台,编写了上位机数据采集、数据存储、数据回放、数据处理等程序模块。本文主要监测主回路电压、主回路电流、线圈电压、线圈电流等参数,从而通过计算或提取波形信号特征等手段来求得接触电阻、吸合时间、燃弧时间、燃弧能量、分断电流等参数。其次,试验中获取的数据包含许多噪声,这些噪声会对状态评估产生干扰,使状态评估不准确,因此需要进行滤波去噪。本文选取具有良好的非平稳性与多分辨分析能力的小波变换进行去噪处理,在小波变换中,阈值和小波层数的选取要非常谨慎,因为它们两个选取的好坏直接关系到去噪的效果。这样经过小波变换等手段进行预处理后,提取信号特征,获得相应特征参数。最后,基于经过分析和预处理的数据,运用灰色模糊综合评判对交流接触器状态进行评估,此方法充分的考虑了评估中存在的灰色性、模糊性、随机性,使评估更加准确。此方法最重要的就是权重的计算和灰色模糊判别矩阵的生成,其灰部都是利用专家打分获得,权重的模部是通过层次分析法确定的,灰色模糊判别矩阵的模部是由三角隶属函数得到的。
[Abstract]:The AC contactor, as a switchgear, is on and off more frequently. At the same time, it can also be used as a control appliance. It can also control electric loads such as electric motors, electric heaters, welding machines and electric power units. It is widely used in the low voltage distribution system and is a very important low voltage electrical apparatus. Therefore, its safe and stable operation is also very necessary for the safe and stable operation of the power grid, in order to ensure its safe and stable operation, Is to be able to monitor their status in real time and to make timely and effective response. In recent years, with the maturation of online monitoring technology, it is more and more popular to evaluate the state of electrical appliances based on big data. Therefore, this paper takes CJX2-80 as an example to study the design of test platform, feature parameter extraction and denoising, state evaluation and so on. First of all, the whole life test platform of AC contactor is designed and built. The test platform includes coil drive circuit, signal acquisition circuit, signal conditioning circuit and so on. The software of the test system adopts the LabVIEW graphic programming language for modular design. It is a virtual instrument development platform provided by the National instrument Company of the United States. The software has compiled the upper computer data acquisition, data storage and data playback. Data processing and other program modules. In this paper, the main circuit voltage, main loop current, coil voltage, coil current and other parameters are monitored, and the contact resistance, absorption time, arc energy and so on are obtained by calculating or extracting the characteristics of the waveform signal. Breaking current and other parameters. Secondly, the data obtained in the experiment contain a lot of noise, which will interfere with the state evaluation and make the state evaluation inaccurate, so it is necessary to carry out filtering and de-noising. In this paper, wavelet transform with good non-stationary and multi-resolution analysis ability is selected for denoising processing. In wavelet transform, the selection of threshold and wavelet layer number should be very careful. Because of their two selection of good or bad directly related to the effect of denoising. After pretreatment by wavelet transform, the signal features are extracted and the corresponding feature parameters are obtained. Finally, based on the data analyzed and preprocessed, the state of AC contactor is evaluated by grey fuzzy comprehensive evaluation. This method fully considers the grey, fuzzy and randomness in the evaluation, which makes the evaluation more accurate. The most important part of this method is the calculation of weight and the generation of grey fuzzy discriminant matrix. The grey part is obtained by expert scoring, and the module part of weight is determined by analytic hierarchy process (AHP). The module of grey fuzzy discriminant matrix is obtained from triangular membership function.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM572.2

【参考文献】