Bit-string optimization in a brachytherapy dosimetry problem

A problem in radiotherapy planning is solved by applying the following search algorithms: artificial genetic search, with fixed crossover probability, with variable crossover probability, and with cut-point distribution; simulated annealing, serial and parallel; stochastic hill-climbing; next-ascent hill-climbing; and steepest-ascent hill-climbing. Given a number of points at each of which a predetermined dose of radiation is required, the problem is to optimize the settings of the radiation sources present in each of three directions (X, Y, and Z) in a three-dimensional selection device such that the sum-squared dose variation is minimized.<<ETX>>