| 基于IAGA-BP神经网络改进的降水测量 |
 点此下载全文 |
| 引用本文：蔡辉，李寅，唐乃乔.基于IAGA-BP神经网络改进的降水测量[J].计算技术与自动化,2025,(3):128-132 |
| 摘要点击次数: 280 |
| 全文下载次数: 191 |
|
|
| 中文摘要:针对降水的小型化的精准测量的需求，采用了多元压电陶瓷片采集降水信号作为神经网络模型输入，并且对AGA-BP神经网络算法进行改进。首先使用该雨量计和标准雨量桶同时采集降水数据获得后期建模所需数据样本，再使用GA-BP神经网络对数据样本进行建模。融入自适应遗传算法对模型进行优化后，又针对传统AGA-BP神经网络进化后期个体适应度差异减小选择过程弱的缺点，使用了非线性函数对变异和交叉因子进行自适应调节。实验结果与数据分析显示：使用IAGA-BP神经网络拟合的降水模型的回归系数高达0.96，MSE为0.014627，决定系数R2为0.88，精度得到明显提升，为全自动化测量降水优化手段提供了新的参考。 |
| 中文关键词:降水 压电陶瓷 神经网络 自适应遗传算法 |
| |
| Improved Precipitation Measurement Based on IAGA-BP Neural Network |
|
|
| Abstract:In order to meet the demand of miniaturized and accurate measurement of precipitation, this study uses multi-component piezoelectric ceramic pieces to collect precipitation signal as the input of neural network model, and improves the AGA-BP neural network algorithm. Firstly, the rain gauge and the standard rain barrel were used to collect precipitation data at the same time to obtain the data samples required for later modeling, and then GA-BP neural network was used to model the data samples, and adaptive genetic algorithm was incorporated to optimize the model, and the weak selection process was reduced due to the individual fitness difference in the late evolution of the traditional AGA-BP neural network. The nonlinear function is used to adjust the variation and crossover factors.Experimental results and data analysis show that the regression coefficient of the precipitation model fitted with the IAGA-BP neural network is as high as 0.96, the MSE is 0.014627, and the determination coefficient R2 is 0.88, which significantly improves the accuracy and provides a new reference for the fully automated precipitation measurement optimization method. |
| keywords:precipitation piezoelectric ceramics neural network adaptive genetic algorith |
| 查看全文 查看/发表评论 下载pdf阅读器 |
