Abstract In recent years, with the development of national economy and the continuous improvement of the electrical level of daily life, various household appliances have gradually entered the field of life. Small circuit breakers have become the most widely used terminal protection equipment in electrical terminal distribution equipment of buildings, its performance plays a key role in the safety of circuits and electrical appliances. Although the size of the small circuit breaker is small, its working principle is not simple. Especially the short circuit breaking of miniature circuit breaker is very complicated. However, most of the faults of small circuit breakers are mechanical ones. The vibration signal of the circuit breaker during closing and closing contains its state characteristic information, the vibration signals of different faults will have certain differences. This paper studies the fault characteristics of small circuit breaker engine, combining fractal technology and probabilistic neural network, a research method of fault diagnosis model for small circuit breaker is proposed. In this paper, the characteristics of probabilistic neural network (PNN) are analyzed in detail, and the fault diagnosis method of probabilistic neural network is used to train samples and classify the faults accurately. Probabilistic neural network is an effective method for fault identification of small circuit breakers. A sample training library was established to facilitate fault classification. In order to classify and identify faults, a fault diagnosis method for miniature circuit breaker based on fractal technique and probabilistic neural network is adopted. The feasibility of this method is verified by data analysis.

中文翻译：

---

基于分形理论和概率神经网络的自适应微型断路器故障诊断模型

摘要 近年来，随着国民经济的发展和日常生活用电水平的不断提高，各种家用电器逐渐进入生活领域。小型断路器已成为建筑物电气终端配电设备中应用最广泛的终端保护设备，其性能对电路和电器的安全起着至关重要的作用。小型断路器的体积虽小，但其工作原理并不简单。特别是微型断路器的短路分断非常复杂。但是，小型断路器的故障大多是机械故障。断路器合、合闸时的振动信号包含其状态特征信息，不同故障的振动信号会有一定的差异。本文研究了小型断路器发动机的故障特征，结合分形技术和概率神经网络，提出了一种小型断路器故障诊断模型的研究方法。本文详细分析了概率神经网络(PNN)的特点，采用概率神经网络的故障诊断方法对样本进行训练，对故障进行准确分类。概率神经网络是一种有效的小型断路器故障识别方法。建立样本训练库，方便故障分类。为了分类和识别故障，采用基于分形技术和概率神经网络的微型断路器故障诊断方法。通过数据分析验证了该方法的可行性。