| The paint drying room is an important part of the automobile production line in the automobile factory,the purpose is to use the heat generated by the combustion of natural gas to bake the car body to cure the paint or coating into a paint film.While the paint drying room consumes a large amount of natural gas,there is also a certain degree of natural gas waste.In today’s increasingly important energy issues,under the new situation of achieving carbon neutrality goals,it is of great practical significance to carry out research on optimization and energy saving of paint drying rooms.The traditional control method has certain limitations in saving natural gas.In this thesis,an intelligent algorithm is used to analyze the real production data and establish a prediction and optimization model in order to achieve a good optimization and energy saving effect.This research takes the intelligent energy-saving project of a factory of a domestic automobile company as the background,and obtains the production data of the paint baking room of the factory,including the data of natural gas consumption,supply air and return air temperature in each area,and the opening degree of the control valve in each area.Based on the preprocessed data set,predictive and optimization models were built and used to achieve energy saving goals.The main work of this thesis is as follows:(1)Based on the LSTM(Long Short-Term Memory)algorithm,an oven temperature prediction model was established,and the hyperparameters of the model were optimized using the PSO algorithm.After verification,the temperature prediction model can accurately predict the temperature of the drying room,and the prediction accuracy of the model after hyperparameter optimization is higher.(2)Based on the GRU(Gated Recurrent Unit)algorithm,the energy consumption prediction model of the drying room was established,and the hyperparameters of the model were also optimized.After verification,the model after hyperparameter optimization can accurately predict the energy consumption of the drying room,which proves that the two prediction models can accurately map the nonlinear relationship between the relevant variables.(3)The GA(Genetic Algorithm)and PSO(Particle Swarm Optimization)algorithm are mixed in a certain way into a GA-PSO algorithm to combine the advantages of the two,and this algorithm is used to establish a temperature prediction model based on LSTM The model is an optimization model of drying room energy consumption with operating constraints and GRU energy consumption prediction model as the evaluation function.After verification,the GA-PSO energy consumption optimization model can find the operation strategy that minimizes natural gas consumption while meeting production requirements.The simulation results show that the optimization algorithm used in this thesis can achieve6.9% savings in natural gas consumption in the paint drying room.It is proved that this research can not only reduce the cost of enterprises and increase the income of enterprises,but also help to achieve the national goal of energy conservation and emission reduction,which is of practical significance. |
