Study On Powered Descent And Hazard Avoidance Guidance Algorithms For Planetary Soft Landing

Powered descent and hazard avoidance guidance is one of the key technologies of safe planetary soft landing, which directly determines the success or not of the entire mission. In addition to index of fuel optimality, the capability of touching down on rough terrain almost unknown in ad vance autonomously and safely at any time is also required for future lunar and martain landing probe. Therefore, it is critical to improve the capability of Chang’e-3 lunar landing guidance scheme to adapt any light conditions and even low visibility.In this thesis, innovative powered descent and hazard avoidance guidance methods for lunar soft landing is studied from the following three aspects: soft landing guidance scheme, hazard detection and optimal landing site selection strategy, guidance algor ithm performance comparison and error analysis. Firstly, Chang’e-3 landing guidance, looked as a typical scheme, is described and analyzed in detail. In order to overcome the shortages of Chang’e-3 landing guidance, an improved candidate scheme is proposed based on the developing 3D imaging flash lidar. According to the given lunar environment models, dynamics models of each sub-phase during powered descent and landing hazard avoidance process are established. Secondly, the guidance algorithms of each sub-phase, and hazard detection and safe landing site selection methods adopted in the typical scheme are summarized and analyzed by computer simulation. The simulation results are almost the same with real flight results of Chang’e-3. Thirdly, minimal fuel nominal trajectory and optimal feedback tracking guidance for powered descent, as well as minimal fuel landing avoidance guidance based on improved hazard detection and landing s ite selection strategy are designed respectively using Gauss pseudospectral method. The effectiveness and superior landing avoidance ability of the improved scheme are verified by mathematical simulation. Finally, the guidance performance of the two schemes mentioned above is quantitatively compared, and the influence of model error on guidance accuracy is quantitatively analyzed. The initial error propagation rules are analyzed by Monte Car lo method and linear covariance analysis method. The results show that the improved guidance laws outperform the typical guidance laws in both fuel consumption and guidance accuracy under the same conditions and that the effect of random initial error can also be effectively suppressed during the guidance process.