Research Based on Improved CNN-LSTM Water Quality Prediction Model

Water quality prediction is an important way to protect water sample area. Most Marine environmental pollution comes from the pollution of shore-based water quality, so the monitoring and prediction of shore-based surface water cannot be ignored. In this paper, the parameters of the model are optimized based on the common water quality prediction models convolutional neural network (CNN) and long short-term memory network (LSTM). First, the basic LSTM neural network prediction model is built, and then the CNN model is integrated to make the prediction effect more close to the real value. Secondly, the swarm intelligence algorithm is used to optimize the parameters of the constructed PHz value and ammonia nitrogen prediction model. Finally, the swarm intelligence algorithms such as firefly algorithm(FA), sparrow search algorithm (SSA) and particle swarm optimization algorithm (PSO) are compared. The SSA-CNN-LSTM neural network prediction model is superior to PSO-CNN-LSTM and FA-CNN-LSTM neural network prediction model, and the SSA-CNN-LSTM prediction model has shorter optimization time and better prediction effect.