Design And Implementation Of Air-Conditioning Load Predicting Algorithm Based On Support Vector Machine

With the increasing proportion of power consumption for air-conditioning in whole city,a smart strategy for air-conditioning control is highly needed.Therefore,in order to have a better control of air-conditioning and to decrease the power consumption of air-conditioning,an air-conditioning load prediction is highly recommended.In other words,if an air-conditioning's load could be predicted in advance then a better strategy to control the air-conditioning could be produced in according to the load.Therefore,turning on the air-conditioning as necessary will result in the decrease of related expenditure as well as the power consumption.In terms of air-conditioning control strategy,air-conditioning control system often apply prediction and then control.In other words,the system will predict the real-time load and then turn on or turn off the air-conditioner according,to the load prediction,in this way,the system will try to running the air-conditioner in daytime while the cost for power is higher as less as possible.Therefore,the power consumption and cost will be decreased.Air-conditioning load prediction has a large impact on air-conditioning's control strategy,and this control strategy has large impact on the consumption of an airconditioning.In order to have a more ecofriendly usage of air-conditioner,an airconditioning's load need to be accurately predicted.Air-conditioning Load Prediction plays an important part in air-conditioning system.A good air-conditioning load prediction system is able to correctly predict the load of an air-conditioning,accordingly,the temperature control of the air-conditioning will be adjusted.The accuracy of the predicting system decides whether the air-conditioning could reach the best temperature control as well as operating under the minimal consumption.Load predicting is vital for an air-conditioning to save energy.Based on the load predicting of air-conditioning,this paper improved an algorithm for predicting by using SVM(support vector machine).In the past research,algorithm for predicting air-conditionings' load often come with a low accuracy,the improved algorithm proposed in this paper comes with a better accuracy and generalization.A system for air-conditioning load prediction based on SVM was designed,and the accuracy of the system was verified.Basing on the current algorithm for load prediction,this paper have a thorough study in terms of those algorithms,and the results are:(1)For air-conditioning's load prediction,this paper proposed a new algorithm basing on SVM.In consideration of the chaotic characteristic of input data,combing with the advantage of SVM that it has a stronger ability to learn,the new algorithm have a Phase Space Reconstruction for input data.Meanwhile,this paper adopted Simulated Annealing to find the best parameters for SVM.Error Correction Model is used for error correct the accuracy of the predicting result.(2)In terms of air conditioning load predicting system,after fulfilling the design of air-conditioning load prediction algorithm,a detailed design was proposed in accordance of the demand analysis,and an air-conditioning load predicting system was implemented based on SVM,last,the quality of the system was verified to satisfied the requirements.The novelties of this paper are as follows: first of all,this paper improved current existing Simulated Annealing algorithm so as to improve the accuracy of predicting.Secondly,this paper combined those methods,Phase Space Reconstruction for input data,parameter tuning for simulated annealing algorithm,ARMA error correction model and support vector(SVM)to implement a support vector machine,and an error correction model is applied to improve the accuracy.After several experiments,the results showed that for predicting air-conditioning's load,the system proposed in this paper have a higher accuracy and generalization.This paper have important meaning for air-conditioning load prediction based on support vector machine for both algorithm design and practical application.