Exploration of dynamic energy management mechanism in cigarette factory based on artificial neural network

In order to solve the problem of excessive reliance on manual experience decision-making in energy management of cigarette factories, a study was conducted on the energy consumption prediction model of cigarette factories. Using a comparative analysis method, two methods, an artificial neural network and multiple regression analysis, were used to construct the energy consumption model of cigarette factories for the same data sample. The energy consumption model was evaluated from three dimensions: coefficient of determination, mean absolute error, and root mean square error. The results showed that the energy consumption model of cigarette factories based on the BP neural network had smaller deviations and better accuracy than the multiple regression analysis method. Furthermore, this article investigates the dynamic energy management mechanism of cigarette factories based on energy consumption prediction models. Dynamic energy consumption prediction is applied to the energy management of the factory, achieving functions such as dynamic prediction, refined assessment, and energy consumption warning, thereby improving the energy management level of cigarette factories.