Scheduling experiments on the space station

The objective of this research is to schedule laboratory experiments on the space station in a power starved environment. Experiments on the space station will compete for electric power and other scarce resources. A principal investigator for an experiment may submit numerical preferences for different periods in which his/her experiment can be started. Weighted preferences are computed by multiplying these preferences by a numerical priority assigned to the experiment by a space station management committee. The scheduling problem is formulated as an integer programming problem with zero/one variables. The objective is to maximize the sum of the weighted preferences for the start periods for all the experiments. The constraints to the scheduling problem ensure that (1) an experiment can begin in at most one period, (2) an experiment of known duration will operate in consecutive periods, and (3) the sum of the average power consumed by all experiments in each period will not exceed the total average power available to the laboratory module in that period. Numerical results are obtained for a small example problem involving two experiments over a planning horizon of ten periods. The problem is solved on an IRM PC using a Fortran matrix generator program followed by a zero/one code obtained from a textbook.