Shengying Zhu

Observability-Based Beacon Configuration Optimization for Mars Entry Navigation

In this paper, a novel observability analysis method of a nonlinear navigation system is proposed, and its application to the optimization of beacon configuration for Mars entry navigation is demonstrated. A Lie algebra-based approach to nonlinear observability is used to compute the observability matrix. The quadratic approximation to the Lie derivatives is then used to recursively calculate the observability matrix efficiently. Next, the inverse of the condition number of the observability matrix is chosen as a metric to evaluate the observability quality. To verify the accuracy and effectiveness of the proposed approach, Mars navigation scenarios with ranging measurements from multiple ground-based beacons are considered in simulation examples, and the optimal beacon configurations are obtained using a genetic algorithm. The extended Kalman filter is finally used to demonstrate the navigation performance. It is concluded that the proposed observability analysis method is able to optimize the beacon con...

Real-time Navigation for Mars Final Approach Using X-ray Pulsars

An accurate knowledge of Mars entry condition is the significant requirement for the successful aerocapture and pinpoint landing. In order to develop the real-time navigation scheme for Mars final approach, the feasibility of navigation embedded X-ray pulsar observations is verified, and the navigation performance is comparatively analyzed. In order to choose optimal navigation pulsars from pulsar candidates, the Fisher information matrix is utilized to evaluate the observability from the estimation theory point of view. The optimal navigation pulsars thus are selected such that the determinant of the Fisher information matrix is maximal. Two navigation scenarios based on the 2012 encounter at Mars of Curiosity spacecraft combining with X-ray pulsar based measurements are then considered to demonstrate the navigation performance. Furthermore, a series of research on the error ellipse in Mars B-plane and the distribution of estimated flight path angle indicate that Xray pulsar based navigation, which may provide more accurate knowledge of Mars entry condition, is a potential navigation scheme for Mars final approach in the future.

Hazard detection and avoidance for planetary landing based on Lyapunov control method

Future planetary landers must be capable of detecting hazards in the landing zone and maneuvering to a new and safe site, for the requirements of the scientific task. This paper presents an autonomous hazard detection and avoidance method based on Lyapunov control method for planetary landing. The terrain of the landing zone is first reconstructed using the feature points of pictures at two different time, and the plane of the landing zone was determined by fitting the terrain elevation data. Then, hazards in the landing zone were identified according to the vitual plane. In order to reduce the potential threats by the hazards, an avoidance control law is designed using Lyapunov function method. The control law can guarantee the landers reach the safe site, simultaneously decrease the landing speed to zero. The results of numerical simulation show that the method is satisfactory for hazards detection and avoidance with assumed environments.

A rapid maneuver path planning method with complex sensor pointing constraints in the attitude space

Large-scale and high-resolution perception is easy to achieve for the physical world, if satellite technology was used in Internet of Things (IoT) in the future. Remote sensing satellite is superior to original method for ground target detection and environmental perception, which could be completed through onboard perception sensors. In the process of detection and perception, satellite needs to frequently perform attitude maneuver in order to meet a variety of task requirements. We have to face with multi-object, multi-sensors constrained maneuver problem. Not only the kinematic and dynamics constraints should been taken into account, but the engineering bounded constraints need to be considered. Moreover, sensor pointing constraints should be elaborated and analyzed reasonably. It is increasingly important how to achieve attitude maneuver in these complex constraints safely and rapidly. Firstly, sensor pointing constraints are translated to quadratic form in order to simplify the representation and computation in the attitude quaternion space. Secondly, we propose an improved RRT planning algorithm for spacecraft, which is able to address a variety of sensor pointing constraints. This algorithm will be used as a global planner, in which the uniformly distributed nodes in the expansion space are randomly sampled and the expanded nodes are screened out based on the comparative evaluation function. Finally, simulation results validate the advantages of the proposed algorithm.

Sequence detection of planetary surface craters from DEM data

The research on identification and recognition of impact craters on planetary surface is focused on how to detect them from background. A novel sequence algorithm is proposed to crater detection that utilizes DEM data instead of images. By investigating the features of ideal craters, several constraints can be developed to extract candidate crater edges from other topographies. Based on the fact that the shape of most craters is approximate to an ellipse, the Least Median Square Ellipse Fitting Method can be used to exclude pseudo-edges, and to reserve the real edges which contain the feature of the crater. The location, orientation and other physical parameters of the crater can be determined by fitting real edges to an ellipse based on Robust Least Square Method. Mathematical simulations are performed with the moon DEM data. The results show that the topography-based crater detection algorithm offers an effective method for identification and characterization of ellipse-like impact craters, and the accuracy is high enough.

Trajectory curvature guidance for Mars landings in hazardous terrains

Focusing on the trajectory curvature, this paper presents an innovative and analytical guidance law for the construction of geometrically convex trajectories. Moving along such trajectories, the lander can increase the probability of a safe landing in hazardous terrains. Initially, the curvature theorems of the powered descent trajectories are developed. In these theorems, the inner relationship between trajectory curvature and lander states is revealed, and the state constraints for a geometrically convex trajectory are derived. Next, the trajectory curvature guidance is developed in an analytical formulation by satisfying the state constraints for a convex trajectory, and the selection of the key guidance parameters is investigated. Finally, the performance of the trajectory curvature guidance is analyzed in detail, illustrating its superior hazard avoidance and the camera’s field of view.

A new algorithm for generation of dispatchable networks with temporal constraints

ABSTRACT In most real-world planning domains, it is important to have flexible executive time against uncertainty and disturbance of environment. Execution time can be propagated from dispatchable Simple Temporal Network (STN), which is obtained from filtering the distance graph of STN. To minimise the execution latency, computation of dispatchable STN should be improved. In the previous works, all-pairs shortest paths (APSP) algorithm is used to compute distance graph of STN, so the redundant constraints should be processed in the following filtering process. In this paper, an improved algorithm named Dispatchabile Partial Path (DPP) algorithm, which uses the P3C algorithm to replace APSP and compute the shortest distance of vertices, is presented. To ensure the correctness of dispatchable STN, triangulation uses vertices ordering sorted by execute time. The filtering process is also simplified by time-sorting vertices. Experimental evidence confirms that the DPP algorithm outperforms previous algorithms based on APSP.

Enriching mission planning approach with state transition graph heuristics for deep space exploration

As to support the mission of Mars exploration in China, automated mission planning is required to enhance security and robustness of deep space probe. Deep space mission planning requires modeling of complex operations constraints and focus on the temporal state transitions of involved subsystems. Also, state transitions are ubiquitous in physical systems, but have been elusive for knowledge description. We introduce a modeling approach to cope with these difficulties that takes state transitions into consideration. The key technique we build on is the notion of extended states and state transition graphs. Furthermore, a heuristics that based on state transition graphs is proposed to avoid redundant work. Finally, we run comprehensive experiments on selected domains and our techniques present an excellent performance.

An action guided constraint satisfaction technique for planning problem

This paper presents an action guided constraint satisfaction technique for planning problem. Different from the standard algorithms which are almost domain independence and cannot reflect the characteristics of the planning progress, we discuss how the action rules in planning act in constraint satisfaction problems. Based on the conclusion, an action directed constraint is proposed to guide the variable selected procedure in constraint satisfaction problems. Through theoretical analysis, this technique is prior an order of magnitude in variable select procedure over the ordinary heuristic technique and can be used in constraint-programmed planning problem generally. With the simulation experiments it shows that the algorithm with action guided constraint can effectively reduce the number of constraint checks during the planning procedure and has a better performance on total running time over the standard version.

A geometric dynamic temporal reasoning method with tags

Temporal reasoning is one of the cognitive capabilities humans involve in communicating with others and everything appears related because of temporal reference. Therefore, in this paper a geometric dynamic temporal reasoning algorithm is proposed to solve the temporal reasoning problem, especially in autonomous planning and scheduling. This method is based on the representation of actions in a two dimensional coordination system. The main advantage of this method over others is that it uses tags to mark new constraints added into the constraint network, which leads the algorithm to deal with pending constraints rather than all constraints. This characteristic makes the algorithm suitable for temporal reasoning, where variables and constraints are always added dynamically. This algorithm can be used not only in intelligent planning, but also computational intelligence, real-time systems, and etc. The results show the efficiency of our algorithm from four cases simulating the planning and scheduling process.