Simulation systems are becoming a widely used tool for planning and staff training in different applications. To support the execution of agents' deployment tasks in virtual simulation systems, this work evaluates the effectiveness of search and optimization algorithms with different nature, namely, the Simulated Annealing and the Tabu Search algorithms, in the resolution of a constraint-based and time-limited optimization problem. Particularly, the modeled simulated application problem regards the spatial positioning (or distribution) of agents in tactical positions of a virtual terrain scenario. The work addresses how to model this problem and to solve it during simulation runtime, where strict processing time limits are considered. Experimental results show that the proposed approach permits to efficiently distribute the agents according to the required formation pattern, both in execution time and agent distribution error, with a significant optimization rate.