基于Inception网络结构的光伏板积灰状态识别方法
张中杰 , 权悦 , 高皓宇 , 庞超 , 苏煜 , 韩浩宇
内蒙古师范大学学报(自然科学版) ›› 2024, Vol. 53 ›› Issue (03) : 264 -271.
基于Inception网络结构的光伏板积灰状态识别方法
Identification of Photovoltaic Panel Ash Accumulation State Based on Inception Network Structure
传统的光伏板积灰状态识别方法识别准确率低、速度慢,因此提出一种基于Inception网络结构的 光伏板积灰状态识别模型。该模型以Inception模块作为主体模块,通过在模型初期添加Stem模块,加大Stem模块卷积核尺寸,从而减少输入数据的维度,增大模型初期的有效感受野,提高模型泛化能力。同时引入Swish-SE轻量级注意力机制,增强模型对不同特征的关注度,提高模型的识别准确率。实验结果表明,所提方法的目标识别率为97.05%,较经典卷积神经网络Inception-V3模型和MobileNet-V2模型分别提高1.64%、5.91%。研究提出的积灰状态识别方法具有参数量少、训练时间短、分类效果好的优势,可以满足光伏电站智能化运维的基本要求,具备较好的实用性。
In light of the low recognition accuracy and slow speed of the traditional photovoltaic panel ash accumulation state recognition method, a photovoltaic panel ash accumulation state recognition model based on Inception network structure is proposed in the paper. The model adopts Inception module as the main module, and increases the size of the convolutional kernel of the Stem module by adding the Stem module at the early stage of the model, so as to reduce the dimensionality of the input data, increase the effective sensory field at the early stage of the model, and improves the generalization ability of the model, and at the same time, it introduces the Swish-SE lightweight attention mechanism, which enhances the model's attention to different features, so as to effectively improve the recognition accuracy rate. The experimental results show that the target recognition rate of the proposed method reaches 97.05%, which is 1.64 and 5.91 percentage points higher than that of the classical convolutional neural network Inception-V3 model and MobileNet-V2 model, respectively. The proposed ash accumulation state recognition method has the advantages of small number of parameters, short training time, and good classification effect, which effectively meet the basic requirements of intelligent operation and maintenance of photovoltaic power plants and are of good practicability.
光伏板 / 积灰状态识别 / Swish-SE / Inception模块
photovoltaic panel / ash accumulation state recognition / Swish-SE / Inception module
| [1] |
曲宏伟,王靖雯.积灰对光伏板输出特性影响理论和试验研究[J]. 太阳能学报, 2018,39(8):2335-2340. |
| [2] |
孟伟君,朴铁军,司德亮, |
| [3] |
|
| [4] |
贾科,顾晨杰,毕天姝, |
| [5] |
张博,黄帅,丛伟伦, |
| [6] |
|
| [7] |
赵波,廖坤,邓春宇, |
| [8] |
孙鹏翔,毕利,王俊杰.基于改进深度残差网络的光伏板积灰程度识别[J]. 计算机应用, 2022,42(12):3733-3739. |
| [9] |
白云瑞,魏霞,黄德启, |
| [10] |
|
| [11] |
任欢,王旭光.注意力机制综述[J]. 计算机应用, 2021,41(S1):1-6. |
| [12] |
董跃华,彭辉林.改进Inception结构的图像分类方法[J]. 软件导刊, 2023,22(2):1672-7800. |
| [13] |
|
| [14] |
|
| [15] |
|
| [16] |
王绍文,宋鹏,谭军, |
| [17] |
陈纪宏,卞荣星,张听雪, |
| [18] |
张天,孙连英,杨琰, |
| [19] |
张洪,盛永健,黄子龙, |
安徽省教育厅自然基金重大资助项目“基于能源互联的矩阵变换器控制技术研究”(2022AH040234)
安徽省大学生创新创业训练计划资助项目“基于嵌入式系统的电机数据监测系统”(S202310879050)
/
| 〈 |
|
〉 |