云模型过渡特性分析及其在绝缘子憎水性等级评价中的应用

Analysis of Cloud Model Transition Characteristics and Its Application in Insulator Hydrophobicity Grade Evaluation

针对经典云模型在绝缘子憎水性相邻等级之间评价比较模糊的问题,通过对经典云模型进行改进,提出一种可改善边缘特征提取能力的过渡云模型。首先,基于确定度在0.5附近的不确定性最大这一特点,结合过渡态理论,在保证论域范围与雾化特性与经典云模型相一致的前提下,通过对确定度0.5以上云滴的舍弃-镜像-拉伸-重新生成等步骤,构造过渡云模型;然后,在此基础上,与经典云模型以及传统的基于图像的复合绝缘子憎水性等级判定方法进行对比试验。结果表明,本模型在保证水珠边缘像素信息不确定性的条件下,相比其他方法能较准确地分辨出交叉处的等级,从而验证了过渡云模型具有概念边缘特征的提取能力,且在绝缘子憎水性等级评价方面具有比较好的应用前景。

In response to the problem of fuzziness of the classic cloud model in evaluating the hydrophobicity of insulators between adjacent grades,through improving the classic cloud model,a transition cloud model that can improve the edge feature extraction ability was proposed.First,based on the characteristic of the maximum uncertainty near 0.5 membership,and the transition state theory,on the premise of ensuring the consistency of the domain scope and atomization characteristics with the classical cloud model,the transition cloud model was constructed through the steps of abandoning-mirroring-stretching-rebuilding of cloud drops above 0.5 membership.Comparative experiments were conducted with classic cloud models and traditional image-based methods for determining the hydrophobicity level of composite insulators.The results show that the proposed algorithm can distinguish the level of intersection more accurately than other methods under the condition of ensuring the uncertainty of water drops of edge pixel information,thereby verifying the ability of the proposed transition cloud model to extract conceptual edge features,and its good application prospect in the evaluation of insulator hydrophobicity grade.