基于SARIMA-IPOA-BiLSTM模型的建筑电力碳排放预测OA

With the continuous growth of global energy consumption,the carbon emission control of the power industry is facing severe challenges.In particular,the carbon emission from natural gas power generation is highly volatile due to multiple factors,which brings great difficulties to the formulation of emission reduction policies and the implementation of carbon trading mechanism.To solve this problem,this study proposes a hybrid forecasting model based on SARIMA-IPOA-BiLSTM:SARIMA model captures linear trends and seasonal characteristics;The Improved Pelican Optimization Algorithm(IPOA)optimizes the hyper parameters of bilstm through cubic chaotic mapping and sine cosine strategy;The bilstm network learns nonlinear patterns in SARIMA residuals.The experimental results show that the model has significant advantages over the traditional method.Compared with the single SARIMA-IPOA model,the root mean square error is reduced by 23.1%,the average absolute error is reduced by 22.1%,and the fitting degree is improved to 95.24%.The research results can provide a scientific basis for the accurate prediction and dynamic control of carbon emissions in the power industry,and provide data support for the operation of carbon trading market and the formulation of emission reduction policies.