An Improved Temperature Prediction Technique for HVAC Units Using Intelligent Algorithms

This paper proposed a methodology to automatically search for the best combination of the Learning Horizon (LH) and Predicting Horizon (PH) with the objective of improving the prediction performance for a temperature prediction technique for HVAC units. Three intelligent algorithms, Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Greedy Algorithm (GRA), are integrated into a temperature prediction process for identifying the parameters of a thermodynamic model. The developed temperature prediction technique was tested, validated, and evaluated in a case study. This case study also compared the prediction performances of the three different optimization algorithms mentioned earlier and explored the impact of the LHs and PHs on the prediction performance. By setting up the selection standards for evaluating the prediction performances of the temperature prediction technique, the most suitable algorithm was then selected along with the best combination of the LH and PH.