Due to the problems of high bit error rate and delay, low bandwidth and limited energy of sensor nodes in underwater acoustic sensor network (UASN), it is particularly important to design a routing protocol with high reliability, strong robustness, low end-to-end delay and high energy efficiency which can flexibly be employed in dynamic network environment. Therefore, a reinforcement learning-based opportunistic routing protocol (RLOR) is proposed in this paper by combining the advantages of opportunistic routing and reinforcement learning algorithm. The RLOR is a kind of distributed routing approach, which comprehensively considers nodes’ peripheral status to select the appropriate relay nodes. Additionally, a recovery mechanism is employed in RLOR to enable the packets to bypass the void area efficiently and continue to forward, which improves the delivery rate of data in some sparse networks. The simulation results show that, compared with other representative underwater routing protocols, the proposed RLOR performs well in end-to-end delay, reliability, energy efficiency and other aspects in underwater dynamic network environments.

Dynamic environments such as urban areas are still challenging for popular visual-inertial odometry (VIO) algorithms. Existing datasets typically fail to capture the dynamic nature of these environments, therefore making it difficult to quantitatively evaluate the robustness of existing VIO methods. To address this issue, we propose three contributions: firstly, we provide the VIODE benchmark, a novel dataset recorded from a simulated UAV that navigates in challenging dynamic environments. The unique feature of the VIODE dataset is the systematic introduction of moving objects into the scenes. It includes three environments, each of which is available in four dynamic levels that progressively add moving objects. The dataset contains synchronized stereo images and IMU data, as well as ground-truth trajectories and instance segmentation masks. Secondly, we compare state-of-the-art VIO algorithms on the VIODE dataset and show that they display substantial performance degradation in highly dynamic scenes. Thirdly, we propose a simple extension for visual localization algorithms that relies on semantic information. Our results show that scene semantics are an effective way to mitigate the adverse effects of dynamic objects on VIO algorithms. Finally, we make the VIODE dataset publicly available at https://github.com/kminoda/VIODE.

We study the problem of synthesizing a policy that maximizes the entropy of a Markov decision process (MDP) subject to a temporal logic constraint. Such a policy minimizes the predictability of the paths it generates, or dually, maximizes the exploration of different paths in an MDP while ensuring the satisfaction of a temporal logic specification. We first show that the maximum entropy of an MDP can be finite, infinite or unbounded. We provide necessary and sufficient conditions under which the maximum entropy of an MDP is finite, infinite or unbounded. We then present an algorithm which is based on a convex optimization problem to synthesize a policy that maximizes the entropy of an MDP. We also show that maximizing the entropy of an MDP is equivalent to maximizing the entropy of the paths that reach a certain set of states in the MDP. Finally, we extend the algorithm to an MDP subject to a temporal logic specification. In numerical examples, we demonstrate the proposed method on different motion planning scenarios and illustrate the relation between the restrictions imposed on the paths by a specification, the maximum entropy, and the predictability of paths.

The aim of this article is to present a novel methodology to determine the correct size of the renewable energy sources (RES) and the battery storage system (BSS) that are needed to effectively convert a conventional residential building into a nearly zero-energy building (nZEB). This is attained by properly considering the long-term history of the weather at the location of the building, the energy consumption of the appliances and the performance of the energy management system (EMS). The above are embedded in a cost function that balances the impact factors which highly affect the size and have considerable effect on the reliability and lifespan of the RES and BSS in a nZEB, i.e., the depreciation of the initial investment, the self-consumption rate, the feed-in rate, the simultaneity in generating and consuming power, and the impacts of the depth-of-discharge and the discharge power on the BSS's state-of-health. Therefore, the correct sizing of RES and BSS is provided by considering both the effectiveness and the economic point of views. Specifically, the optimal solution is attained by minimizing the above cost function through the genetic algorithm technique. The effectiveness, functionality and practicality of the proposed methodology have been validated on a pilot nZEB.

Triangular-current-mode (TCM) modulation guarantees zero-voltage-switching across the mains cycle in AC-DC power converters, eliminating hard-switching with a minor 30% penalty in conduction losses over the conventional continuous current mode (CCM) modulation scheme. TCM-operated converters, however, include a wide variation in both switching frequency and switched current across the mains cycle, complicating an analytical description of the key operating parameters to date. In this work, we derive an analytical description for the semiconductor bridge-leg losses in a TCM AC-DC converter, including the rms current and/or conduction losses, switching frequency, and switching losses. For SiC MOSFETs, we introduce a new loss model for switching losses under zero-voltage-switching, which we call “residual ZVS losses”. These losses include the constant Coss losses found in previous literature but must also add, we find, turn-off losses that occur at high switched currents. The existence and modeling of these turn-off losses are validated on the IMZA65R027M1H 650V SiC MOSFET. The complete loss model is validated on a 2.2kW hardware bridge-leg demonstrator, which achieves a peak 99.6% semiconductor efficiency at full load. The proposed, fully-analytical model predicts bridge-leg losses with only 12% deviation, accurately including residual ZVS losses across load, modulation index, and external gate resistance.