Compressed air flow prediction system based on BO-LSTM neural network model

In today's era, the rapid development of industry has made compressed air an important raw material and power source in many fields. However, the problem of high energy consumption in the air compressor workshop still needs to be solved and optimized. The flow prediction of compressed air is an effective means and important link to solve and optimize the high energy consumption of air compressor workshop and air compressor. In this paper, a prediction model based on long short-term memory neural network (LSTM) is established to predict the flow rate of compressed air. After some analysis and comparison, four indexes of exhaust pressure, exhaust temperature, air compressor operating current and gas tank pressure are selected as input features for training neural network and final prediction model. Finally, using Bayesian optimization to modify and improve the model, the BO-LSTM model established has practical significance for the compressed air flow prediction and energy consumption reduction.