Laser Powder Bed Fusion (L-PBF) is a widespread additive manufacturing process in which metal powder is selectively melted by a laser. The choice of process parameters, such as the laser power, has a decisive influence on the part quality [1]. This work focuses on the use of adaptive process parameters that take into account both the exposure strategy and the part geometry (local heat conduction condition). Pichler [2] shows that adaptive process control can significantly improve the surface quality and geometry freedom of L-PBF parts. To develop this adaptive process control, geometry-dependent rules are used to adapt process parameters. In this work, an alternative simulation-based approach is presented. First, a thermal simulation model is developed that includes the close surroundings (approx. 1 mm) of a single scan vector. The simulation model is based on the finite difference method. Subsequently, a model-based adaptation of the laser power using different simulation parameters (e.g. variation of the powder bulk density) is calculated for a part. The calculated adaptive process parameters are then used to manufacture the part. To evaluate the various simulation parameters, the part density is analyzed using a gas pycnometer. In addition to analyzing the part density, the manufacturing process is recorded by pyrometry in the infrared range. The research by Stajkovic et al. [3] showed that the process emission measured by pyrometry is largely determined by the aspect ratio of the keyhole. Due to this correlation, the measured thermal process emission can be used to evaluate the process stability in relation to the stability of the keyhole. The results show that the use of adaptive, simulation-based process parameters leads to a more homogeneous thermal emission. In addition, an increase in part density of approx. 0.15 % can be determined. The test results illustrate the potential of simulation-based adaptive process control.