基于CNN-Transformer的黄河水质参数并行预测模型

打开文本图片集

关键词：水质参数;并行预测；CNN-Transformer模型；黄河中图分类号：X52；TV882.1 文献标志码：A doi：10.3969/j.issn.1000-1379.2026.03.03

Parallel Prediction Model for Yellow River Water Quality Parameters Based on CNN-Transformer

WANG Chaoliang' ， GUO Rongxing' ， ZHAO Xuezhuan 1，2 ，WANG Jun’， ZHAO Niyuan’， CHEN Jimin³ （1.School of Computer Science，， Zhengzhou University of Aeronautics， Zhengzhou 450046，China; 2.Henan Provincial Industry-University-Research Center for Artificial Intelligence Technology， Zhengzhou 450046，China；3.Information Center，YRCC，Zhengzhou 450004，China）

Abstract：Facedwithteprobleofinsufentaccuracyintraditionalwaterqualityparametrspredictionmetodsenealingwithcomplexnonlinearwaterqualityparameterschange，aCN-Transformer-basedparalelpredictionmodelforwaterqualityparametersnte YelowRiverwaspropoed，asedontheperodicndlearharacteristicsofterqualityparameershangeisodelpdicteissolvedoxygen，peangaateindexmmoiarogenndtotalposporusatteQilipumoitoringsctioofteelowiverfro to2025.ThemodelinputedmonitoringdatainparalelintotheCN（ConvolutioalNeuralNetwork）moduleandtheTansforeodule， respectivelyextractinglocaletailaturesandglobaldynamicfatures，ndsedafullyotedlaertoapthefusdfeaturetoepr dictionresults.ComparingthepredictioperformanceoftheCNN-TransforerodelwithRN（RecurentNuralNetwork），N，LSTM （Longhort-TermMemory），andTransformermodels，theresultsshowthat，comparedwiththeotherfourmodels，theCNN-Transforer model reduces MSE by 3.93% 一 10.96% ，RMSE by 5.82% 9.33% ，MAE by 12.44% 1 14.48% ，and improves R2 by 6.56% 26.65% ，demonstrating the most outstanding performance.

Key words：water quality parameter;parallel prediction； CNN-Transformer model； Yellow River

0 引言

黄河作为中华民族的母亲河，流域水环境保护至关重要，是国家生态安全战略的重要组成部分[1]。（剩余5843字）