| 基于GWO改进神经网络的风致输电杆塔响应计算方法 |
| |
| 引用本文： | 谢从珍,马康,卢伟民,王勇.基于GWO改进神经网络的风致输电杆塔响应计算方法[J].科学技术与工程,2023,23(31):13407-13414. |
| |
| 作者姓名： | 谢从珍 马康 卢伟民 王勇 |
| |
| 作者单位： | 华南理工大学电力学院;广东电网有限责任公司广州供电局 |
| |
| 基金项目： | 国家自然科学基金项目（面上项目，重点项目，重大项目） |
| |
| 摘 要： | 针对基于有限元方法的风致输电杆塔仿真模型存在模型参数设置复杂、响应速度低下等问题，本文提出了一种结合杆塔仿真计算与机器学习的杆塔应力快速计算方法，实现了风致输电杆塔仿真模型的高效运算。本文首先利用有限元仿真分析10~35 m/s风速0~90°风向下输电杆塔的应力结果，将其作为神经网络训练样本；分析杆塔本征参数及气象因素对杆塔应力结果的影响，提出一种基于风速、风向和塔材类型等因素的特征值选取方法；利用GWO优化算法提高神经网络准确度，建立基于GWO-BPNN的杆塔风致应力计算模型，计算速度可达建模仿真的10倍以上。在准确率上，基于数据集划分训练集及验证集，模型准确率在98%以上，将计算结果与实际杆塔受灾情况相比情况一致。本模型可用于输电线路防风预警，致力于输电线路防灾减灾。
|
| 关 键 词： | 风致杆塔响应 BP神经网络 杆塔风荷载 杆塔风险评估 防风预警 |
| 收稿时间： | 2022/12/27 0:00:00 |
| 修稿时间： | 2023/8/8 0:00:00 |
Research on the response calculation method of wind-induced transmission tower based on GWO improved neural network |
| |
| Xie Congzhen,Ma Kang,Lu Weimin,Wang Yong.Research on the response calculation method of wind-induced transmission tower based on GWO improved neural network[J].Science Technology and Engineering,2023,23(31):13407-13414. |
| |
| Authors: | Xie Congzhen Ma Kang Lu Weimin Wang Yong |
| |
| Affiliation: | School of Electric Power Engineering,South China University of Technology; Guangdong Power Grid Co,Ltd Guangzhou Power Supply Bureau |
| |
| Abstract: | Aiming at the problems of complex model parameter setting and low response speed of the wind-induced transmission tower simulation model based on finite element method, a rapid calculation method of tower stress combining tower simulation calculation and machine learning is proposed. This method enables efficient computation of the simulation model of wind-induced transmission towers. Firstly, the FE simulation of 10~35m/s wind speed 0~90° wind downward is carried out to analyze the stress results of the transmission tower, and the results are used as neural network training samples. Then, the influence of intrinsic parameters and meteorological factors on the stress results of the tower is analyzed. A method for selecting eigenvalues based on factors such as wind speed, wind direction and tower type is proposed. The GWO optimization algorithm is used to improve the accuracy of the neural network, and the calculation speed of the GWO-ANN model can reach more than 10 times that of the modeling simulation under the same computing power. In terms of accuracy, the training set and the validation set are divided based on the dataset, and the accuracy of the model is above 98%, and the calculation results are consistent with the actual tower disaster situation. This model can be used for wind protection warnings of transmission lines at different times, and is committed to disaster prevention and mitigation of transmission lines. |
| |
| Keywords: | wind-induced tower response BP neural network tower wind load Pole tower risk assessment wind protection warning |
|
| 点击此处可从《科学技术与工程》浏览原始摘要信息 |
|
点击此处可从《科学技术与工程》下载全文 |
