基于AHP-BP神经网络的配电网绝缘作业评估

Assessment of Insulation Operation in Distribution Network Based on AHP-BP Neural Network

摘要:

针对配电网绝缘化作业评估过程繁琐导致的评估效率不高和复杂度高的问题，对智能综合评估模型的构建进行了研究。采用了层次分析法模型与反向传播神经网络结合的技术和方法，进行了评估指标的量化分析，构建科学的评估体系。为进一步提高评估的准确性和减少主观判断的影响，引入反向传播神经网络来映射层次分析法评估过程中的权重分析和评估，从历史数据中学习并提取潜在的、难以量化的影响因素与权重之间的关系。进行了权重决策阈值的设定分析，实现了配电网绝缘化作业的决策优化。经仿真实验测试，验证了模型的有效性和更低的计算复杂度，满足了高效评估需求。经过复杂度分析，提出方案相较于AHP方法有效的将计算复杂度从 降低至 。

Abstract:

To address the issues of low evaluation efficiency and high complexity arising from cumbersome evaluation processes for distribution network insulation operations, a study was conducted on the construction of an intelligent comprehensive evaluation model. The model integrated the analytic hierarchy process with a back propagation neural network. Quantitative analysis of evaluation indicators was performed to establish a scientific assessment system using analytic hierarchy process. Further, to enhance evaluation accuracy and minimize subjective judgment, a back propagation neural network was introduced to map weight relationships during the analytic hierarchy process evaluation process, learning from historical data to extract latent and non-quantifiable influencing factors. Analysis of weight decision thresholds enabled optimized decision-making for insulation operations. Simulation experiments verified the model’s effectiveness, reduced computational complexity, and confirmed its capability to meet efficient evaluation demands. Complexity analysis demonstrates that the proposed method achieves a reduction in computational complexity from to relative to the analytic hierarchy process.