Neural network carbon emission prediction method based on big data environment and application

The invention provides a neural network carbon emission prediction method based on a big data environment, and the method comprises the steps: S1, obtaining multi-channel carbon emission data through an Internet of Things platform network, enabling the multi-channel emission data to form a corresponding emission data set, and carrying out the normalization processing; the method comprises the following steps of S1, acquiring carbon emission data, S2, measuring and calculating the offset of the carbon emission data, performing dynamic correction processing on the data, and calculating a corresponding mapping transformation function through a TCN time convolutional neural network, and S3, acquiring multi-channel carbon emission data fusion through the mapping transformation function, performing carbon emission prediction interval judgment through a prediction precision model, and uploading the carbon emission prediction interval to an Internet of Things platform network. 本发明提出了一种基于大数据环境下神经网络碳排放预测方法，包括：S1，通过物联平台