Research On Illuminance Modeling And Illumination Optimization Control

The comfort of indoor light environment and lighting energy consumption are important indicators of intelligent buildings.The fast and accurate measurement of illumination distribution and the optimal control of lighting have become the focus of attention.At present,light sensor is commonly used to measure the illumination,which can only obtain the average illumination of the region and cannot achieve accurate illumination distribution.On the other hand,the traditional lighting control optimization algorithm is prone to fall into the local optimal solution in the iterative process,so there is still a large room for improvement of lighting energy-saving control.Aiming at the above problems,an illuminance model based on radial basis function neural network is proposed in this thesis to realize the fast estimation of illuminance distribution.The Genetic-simulated annealing algorithm is applied to lighting optimization control,which can minimize lighting energy consumption under comfortable lighting conditions.The main research contents of this thesis are as follows:(1)The illuminance estimation method by illuminance modeling and network training is proposed,which provides a basis for lighting control and effectively solves the problem of slow convergence in the feedback overlapping process of "measurement of illuminance distribution--optimization dimming ratio" in the optimization control algorithm of multi-lighting equipment.(2)Establishment of artificial illumination model.Firstly,the RBF neural network was used to establish the illuminance model of a single indoor lighting device,and the RBF neural network was trained by the measured illuminance data set,so as to obtain the network weight coefficient.The illuminance distribution of indoor multi-lighting equipment is obtained by the principle of superposition of illuminance to establish the artificial illuminance model of multi-lighting equipment.Thus,fast estimation of illuminance distribution can be achieved.The experimental results show that the relative error is less than 5.00%between the illumination distribution obtained by the artificial light illuminance model established in this thesis and the actual measurement results under different artificial light environments,which can meet the requirements of lighting control.(3)Establishment of natural illumination model.Because the distribution of indoor natural light illuminance is affected by many factors such as the longitude and latitude of the building,the orientation of the building,the size and orientation of the Windows,the weather,the season,the height Angle and the direction Angle of the sun,the modeling of natural light illuminance is difficult.In this thesis,a simplified illuminance modeling method is proposed.On the premise of not considering the weather,season,solar altitude Angle and direction Angle and other factors,the radial basis function neural network is used to establish the natural illuminance distribution model in the interior of a building.The model is modified by the real-time measurement value of the illuminance sensor at a fixed indoor position,so as to obtain the distribution estimation of the illuminance under the conditions of weather,season,solar altitude Angle and direction Angle.Through experimental verification,the above model can achieve a more accurate estimation of the distribution of natural light illuminance,and can meet the needs of lighting control.(4)The lighting optimization control problem was reduced to an optimization problem with constraints from the perspective of indoor personnel comfort lighting needs and energy-saving lighting.The genetic-simulated annealing algorithm is proposed aiming at the problem that the traditional optimization control algorithm is easy to fall into the local optimal solution.The idea of simulated annealing is incorporated into the genetic algorithm,and the fast convergence of the genetic algorithm and the robustness and global of simulated annealing are combined to achieve the fast global optimal solution to the lighting optimization problem,and further improve the energy saving of the lighting control system.(5)An experimental platform based on DALI lighting system was built.The lighting optimization method proposed in this thesis was tested to evaluate the energy-saving effect of the lighting optimization control method.The experimental results show that the energy saving of genetic simulated annealing algorithm is improved by 1.60%and 5.00%,respectively,compared with particle swarm optimization algorithm and genetic algorithm for the environment with only artificial light source.When natural light source and artificial light source exist at the same time,the energy saving of genetic simulated annealing algorithm is improved by 5.38%and 8.32%respectively compared with particle swarm optimization algorithm and genetic algorithm,which effectively realizes the energy saving lighting under the premise of comfort lighting.