针对变压器在故障诊断时复杂难辨的问题,提出了一种基于人工鱼群算法优化BP神经网络的故障诊断方法。该方法首先将所得的样本进行预处理,然后通过归一化后的故障特征量样本和目标期望输出,对建立的模型进行样本训练,最后将验证数据进行诊断测试。诊断结果表明,采用该方法可以满足变压器故障诊断的需要,具有很强的实用价值。 Aiming at the problem that the transformer is complex and difficult to distinguish during fault diagnosis, a fault diagnosis method based on artificial fish school algorithm to optimize BP neural network is proposed. The method first preprocesses the obtained samples, and then, samples the established model through the normalized fault feature samples and the expected output of the target. Finally, the verification data is used for the diagnostic test. The diagnosis results show that this method can meet the needs of transformer fault diagnosis and has strong practical value.