As the pillar of China's economic and social development, the industrial sector is not only a major “energy consumer” but also a significant “carbon emitter”. Industrial parks, serving as the main platforms for urban economic activities and growth engines, contribute to over half of the country's industrial output value and account for more than 30% of carbon emissions. This paper explores a method for calculating carbon emissions in industrial parks based on deep learning algorithms for intelligent carbon emissions, clarifies the boundaries for such calculations, adopts the LSTM neural network model, and optimizes it using the FA algorithm to establish a carbon emission calculation model for China's industrial parks. Taking an industrial park in Suzhou as an example, the paper collects relevant data, inputs them into the model for training, and calculates the park's carbon emissions from the 2023–2035 based on the predicted input data. By comparing the calculation results with the park's carbon peaking action plan, the accuracy of the model's calculations is verified.