Abstract
This research deals with physically-based modeling of power system loads. Residential air conditioner system has been chosen to demonstrate the modeling technique. The dynamic behaviour of the air conditioning system is modeled from fundamental physical principles of heat transfer. With certain assumptions the model is simplified and expressions for rate of switching on and off of the air conditioner are obtained. The parameters of the model are estimated using system identification technique. A scheme for Maximum Likelihood identification of linear dynamic systems is developed. Remedies to the problems associated with identifiability are presented and have been successfully used to identify the parameters of the air conditioning model. The algorithm is tested under various simulation conditions. A sample of houses is selected with some of the parameters different from the test case. The parameters of these houses are identified. The identified models are than validated by comparing the estimated values of the duty factors obtained using the identified models and the values of the duty factors obtained by simulation. The mean and variance of the duty factor are estimated and compared with the simulated values to further assess the validity of the identified models. The model is then used to study the energy consumption of the air conditioners, which is computed using the duty factors. A test case is chosen and the parameters are varied one at a time. The energy consumption for all the cases is determined and the correlation between change in the parameter value and energy consumption is studied. Next, the effect of voltage variation on the performance of the air conditioners is determined. A voltage dependent model of the single-phase induction motor is developed. This model determines the effect on the power output of the motors for change in voltage. The motor model and the air conditioner model are interfaced to obtain the characteristics of the air conditioners as function of voltage. Finally, an evaluation of the physically-based modeling methodology is presented and suggestions on future work to further enhance the scope of these methods are presented.
Pahwa, Ani (1983). Physical-stochastic modeling of power system loads : modeling and system identification of residential air conditioning system. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -582329.