Application Of Support Vector Machine In Street LAMP Energy Saving Control Strategy

The traditional street lamp industry mainly uses time,longitude and latitude,illumination and other strategies to control the street lamp switch,and uses interval lighting,timing power adjustment to adapt to the changing traffic flow.On the one hand,in the street lamp switch method,the theoretical energy-saving effect of illumination control is the best,but the energy-saving effect is not maximized because of the influence of environmental factors such as data acquisition error and installation angle.On the other hand,in the power regulation method,the timing regulation power is relatively good,but the timing control can not adapt to the changing situation of the traffic flow in all seasons and holidays,so there is room for further improvement.In view of the shortcomings of the existing street lamp switch control and power control,this paper proposes a new energy-saving control strategy for street lamp based on clustering algorithm and support vector machine algorithm.Through the analysis of different index data,the time and power of street lamp are worked out to achieve the purpose of effective energy-saving.Experiments on accuracy and recall rate prove that this strategy is effective.The main research work completed in this paper is as follows:(1)To solve the problem of unstable lighting switch control,this paper proposes a street lamp switch control method,SVM-ES,which combines illumination clustering and support vector classifier algorithm.This method collects the illumination,time and installation angle data,and uses K-means algorithm to cluster the illumination data.It changes the illumination data to five levels(1-5).Then it trains the data through SVM to predict the switching time of street lamps without considering other external factors.The experimental results show that the algorithm can effectively reduce the power consumption of street lamps.(2)Aiming at the problem that the timing power regulation can not adapt to the change of traffic flow,this paper proposes a street lamp power regulation method,SVR-ES,which integrates traffic flow clustering and support vector regression algorithm.Firstly,the data of traffic flow,time and road type are collected,and K-means algorithm is used to cluster the traffic flow data,which changes the original traffic flow data to five grades(1-5).Then,the data are trained by SVR to predict the power percentage of street lamp without considering other external factors.Ratio.The experimental results show that the algorithm can effectively reduce the power consumption of street lamps.(3)SVM-ES and SVR-ES study street lamp energy-saving from different perspectives,so they can be combined,and a new method of street lamp energy-saving based on support vector machine classification and regression is proposed.Firstly,the illumination,time and installation angle data are collected,and the illumination data are clustered and grouped by K-means algorithm.Then,the data are trained by SVM to predict the switching time of street lights.Then,the traffic flow,time and road type data are collected,and the traffic flow data are aggregated by K-means algorithm.Class I grouped the data by SVR to predict the power percentage of the street lamp in the time of turning on the lamp.The experimental results show that the proposed algorithm can reduce the power consumption of street lamps more effectively than the two methods mentioned above,namely,SVM-ES and SVR-ES.(4)According to method three,a prototype system is established based on the method of street lamp energy saving based on support vector machine classification and regression.The prototype system is relatively independent from the street lamp control system and provides the interface of the HTTP protocol.The street lamp control system transmits the data of illumination,traffic flow,time,installation angle and road type to the prototype prediction system.Then the prototype prediction system calculates the results of the switch lamp status and power percentage and returns them to the street lamp control system.