For a research facility, the capacity to control and simulate the widest range of operating conditions is key to evaluate and qualify the hosted processes. Work is conducted for solar tower plants (STP) in order to develop heliostat aiming strategy to tailor flux distribution for the targeted needs, either in simulation, experimental or operational scale, but nearly none for solar furnaces with their double reflection, despite their capability to investigate a much larger flux range. The work presented here, funded by the European ERDF project CLÉ-DE-SOL, has allowed to greatly enhance flux control for Odeillo’s Big Solar Furnace operated by the PROMES-CNRS laboratory: both spatial and dynamic control of incoming power on the tested solar receivers. Firstly, flux measurement methods and heliostats upgrades have allowed to build an accurate heliostats flux database. Secondly, we have developed an optimization algorithm based on genetic heuristics in order to determine where each heliostat should aim, based on the flux distribution from the heliostats flux database, in order to reach together the required flux distribution, despite the double reflection (heliostat+parabola) and the cavity of the test case which greatly complicate the problem compared to typical approaches for solar towers. The method presented here should also be applicable for solar towers with secondary concentrator and/or deep cavity for the solar receiver.