Robust pose graph optimization is essential for reliable pose estimation in Simultaneous Localization and Mapping (SLAM) system. Due to the nature of loop closures, even one spurious measurement could trick the SLAM estimator and severely distort the mapping results. Existing methods to avoid this problem mostly focus on ensuring local measurement consistency by evaluating measurements independently, often requiring parameters that are difficult to tune. This paper proposes a cluster-based penalty scaling (CPS) method to ensure both the local and global consistency by first evaluating the edge quality locally, and then integrating this information into the optimization formulation.

Due to their efficient locomotion and natural tolerance to hazardous environments, spherical robots have wide applications in security surveillance, exploration of unknown territory and emergency response. Numerous studies have been conducted on the driving mechanism, motion planning and trajectory tracking methods of spherical robots, yet very limited studies have been conducted regarding the obstacle avoidance capability of spherical robots. Most of the existing spherical robots rely on the “hit and run” technique, which has been argued to be a reasonable strategy because spherical robots have an inherent ability to recover from collisions. Without protruding components, they will not become stuck and can simply roll back after running into bstacles. However, for small scale spherical robots that contain sensitive surveillance sensors and cannot afford to utilize heavy protective shells, the absence of obstacle avoidance solutions would leave the robot at the mercy of potentially dangerous obstacles. In this paper, a compact magnetic field-based obstacle detection and avoidance system has been developed for miniature spherical robots. It utilizes a passive magnetic field so that the system is both compact and power efficient. The proposed system can detect not only the presence, but also the approaching direction of a ferromagnetic obstacle, therefore, an intelligent avoidance behavior can be generated by adapting the trajectory tracking method with the detection information. Design optimization is conducted to enhance the obstacle detection performance and detailed avoidance strategies are devised. Experimental results are also presented for validation purposes.

We propose a novel formulation to use plane primitives in the direct sparse odometry formulation. Unlike existing SLAM works that introduce geometric error terms to account for planes, our method keeps the full error as purely photometric. The proposed system exploits the segmentation masks and plane parameters from a deep neural network and jointly optimizes the camera poses, 3D points, and 3D planes. The tightly-coupled formulation improves the pose estimation by fusing information on high-level geometric primitives and also refines the outputs of the plane detection network during optimization.

Aiming at the problems of slow dynamic response and poor robustness in the traditional double closed-loop PI control method of interleaved parallel DC/DC converters, an improved double closed-loop control method based on Linear active disturbance rejection control (LADRC) is proposed. In this method, the outer loop voltage controller is built by the Reduced order extended state observer (RESO), an adaptive duty cycle compensation control strategy is designed according to the system output voltage feedback error, and the Gray wolf optimization algorithm (GWO) optimizes the parameters of the outer loop controller. Meanwhile, a simple penalty mechanism is proposed to improve the “falling” phenomenon of GWO that is easy to overshoot the parameters of the outer loop controller. Compared with the double closed-loop PI and the second-order LADRC, the simulation results show that the proposed method can achieve a more superior tracking response in the start-up phase, and has a smaller voltage drop and faster recovery time in the case of a sudden load change, which effectively alleviates mutual restriction between dynamic response speed and robustness.