Optimization of Four-Limit Practical Grid-Type LED Layout in Visible Light Communication System Based on K-Means-SCSO-RBF

Abstract

To ensure the uniform signal distribution of indoor visible light communication system and realize forecasting the optimal light source layout scheme under the random room state, this paper proposes a four-limit practical grid-type light source layout scheme that integrates the sand cat swarm algorithm and the RBF neural network of K-means clustering to realize the optimal design of the light source layout. Considering one reflection from the wall, the room state data and the actual optimal position coordinates of LEDs are used as the training dataset utilized to train the K-means-SCSO-RBF neural network model. The optimal indoor light source layout prediction model is established. The simulation results indicate that the model’s average prediction error for 20 randomly selected room states is 0.0151 m. The prediction errors for the 80 selected room states are mainly centered within 0 m to 0.01 m. Therefore, this study aids in identifying the optimal room light source layout. Therefore, the research content of this paper helps to determine the optimal layout of visible light sources in any room. It has the advantages of small prediction error, practicality and generalization. It provides favorable theoretical support for the layout of indoor visible light sources.