Aiming at the limitations of traditional visible light positioning methods, such as complex spatial layout and high hardware costs, this paper proposes an adaptive visible light positioning virtual simulation system based on the difference in received signal strength. By improving the Lambertian radiation model's optical channel gain formula, a nonlinear mapping relationship between light intensity and distance is derived. The system innovatively integrates the local mean decomposition denoising algorithm with the adaptive extended Kalman filter algorithm. Simulation results show that within a positioning range of 0.7m, the system's average positioning error is 1.038 cm, which is 60.1% more accurate than the trilateration method used in reference 10. The innovation of the experiment lies in the fact that through algorithm optimization, the traditional trilateration method's three-sensor scheme is simplified to a combination of two sensors and a single servo. This not only reduces the dependence on multi-node spatial layout but also provides a more cost-effective technical solution for small-range high-precision passive positioning.
ZHU et al. (Sun,) studied this question.