Design of Indoor Induction Lighting for Smart Hotels Based on Bayesian Optimization Gradient Boosting Regression Tree
A method for predicting indoor sunlight illuminance based on Bayesian optimization gradient boosting regression tree is proposed to address the problem of non-linear changes in natural light that prevent hotel environments from maintaining good illuminance at all times.This method uses a Gaussian function to predict the peak illuminance generated by a single lighting device,and then intro-duces a Gaussian mixture error model to compensate for prediction errors.Finally,the Bayesian optimi-zation gradient boosting regression tree is used to predict indoor sunlight illuminance.The experimental results show that under different numbers of personnel,the prediction error values of the Bayesian opti-mized gradient boosting regression tree prediction model are 32 Lux,47 Lux,31 Lux,and 17 Lux.The proposed prediction model can accurately predict indoor sunlight illumination,thereby achieving intelli-gent dynamic control of induced lighting,ensuring the efficient operation of indoor lighting systems,and providing decision-making reference for induced lighting control in lighting systems.
万方数据
维普
