A direct and relatively simple method to cut down computational runtime in forecast software by up to 86.

Area of Science:

• Computational engineering
• Building energy simulation
• Software optimization

Background:

• Software runtime is a major challenge in industries requiring complex simulations, particularly for HVAC forecasting.
• Repeated simulations for intricate input requirements in HVAC, cooling, and heating applications are time-consuming.

Purpose of the Study:

• To present a Python software suite that minimizes mathematical operations and streamlines iterations for cumulative output profiles.
• To significantly reduce simulation runtime for HVAC energy consumption forecasting.

Main Methods:

• Developed a Python software suite to optimize simulation processes.
• Utilized a "representative day" from a short reference year with reduced input files.
• Employed the PATE method to adjust errors between simulated and real software-generated results when deviations were significant.

Main Results:

• Achieved an impressive reduction in simulation runtime, up to 86%.
• Demonstrated the effectiveness of a problem-specific algorithm for HVAC energy consumption.
• Successfully applied the method to diverse buildings across multiple continents.

Conclusions:

• Simple, problem-specific algorithms can yield significant improvements in simulation/CAD runtime.
• The developed method offers an efficient approach to reduce computational load in forecasting applications.
• The software suite is applicable to various buildings and geographical locations, highlighting its versatility.