In view of the difficulty of existing air conditioning control methods in compromising load reduction and heating comfort objectives by smaller group numbers with simple control themes

a new method called clustering and peak-staggered(CPS)control is proposed. To begin with

two thermal parameters

temperature variation exponent and characteristic temperature difference

are used as characteristic attributes to group the clusters at the controlled ends. To maintain status diversity of the controlled ends while avoiding load oscillation during the peak clipping period

the concept of standby time is introduced and a minute load minimization problem is proposed for optimizing the controlled ends' sequence of gearing into the load reduction operating mode. A numerical example is provided to reveal the characteristics and load regulation mechanism of the CPS control method

and to compare the method presented with various exiting ones from the aspects of load regulation performance

thermal comfort effect and load waveforms. Results show that the CPS control method

by staggering the load reduction beginning time of the controlled ends

can achieve rapid air conditioning load reduction while satisfying thermal comfort effect by smaller grouping numbers.