Predicting Nitrogen Oxide Concentration Based on Quadrupedal Robots

Over the past few decades, the field of environmental atmospheric pollution prediction has undergone relentless exploration and progress, with significant advancements in predictive methodologies and technical measures. However, further improving the accuracy of air pollution predictions and flexibly predicting them remains at the core of this field. To address the issues of prediction accuracy and flexible forecasting for air pollutants, this paper proposes a nitrogen compound concentration prediction model based on a Genetic Algorithm and Particle Swarm Optimization optimized Back Propagation (BP) neural network, deployed on a quadruped robot platform. Under the framework of the BP neural network, this model integrates the Genetic Algorithm and Particle Swarm Optimization to form a hybrid optimization method. This method can optimize the initial weights of the BP neural network, thereby significantly enhancing the generalization performance of the neural network and preventing the network from prematurely converging to local optimal solutions. The model is used to predict the concentration of NO2, and it is also compared with other models through experiments. The results of the prediction and comparative experiments show that this prediction model demonstrates high prediction accuracy. Its prediction results outperform other models, achieving the expected effect.