System stability hinges on the imposition of restrictions on the quantity and distribution of deadlines that are missed. The formal articulation of these limitations is as weakly hard real-time constraints. The current state of research in weakly hard real-time task scheduling involves the construction of scheduling algorithms. These algorithms are intended to provide guarantees regarding constraint fulfillment, while seeking to maximize the total quantity of timely task completions. pacemaker-associated infection Within this paper, a significant literature review investigates the connections between weakly hard real-time systems and the design of control systems. The description of the weakly hard real-time system model, including the scheduling problem, is offered. Further, an exploration of system models, constructed from the generalized weakly hard real-time system model, is provided, with a specific emphasis on models applied to real-time control systems. A comprehensive review and comparison of the state-of-the-art algorithms for scheduling tasks constrained by weak real-time deadlines is conducted. To conclude, this section details strategies for controller design built upon the weakly hard real-time framework.
For Earth observation tasks, low-Earth orbit (LEO) satellites necessitate attitude adjustments, which are broadly categorized into two types: maintaining a specific orientation towards a target and shifting between different target-oriented positions. The former's determination rests on the observed target, but the latter, with its nonlinear nature, necessitates careful consideration of various contributing factors. Henceforth, developing an optimum reference posture profile is a complex endeavor. Not only are mission performance and ground communication of the satellite's antenna determined by the target-pointing attitudes, but these are also reliant on the maneuver profile. To improve observation image quality, maximize achievable mission counts, and boost the accuracy of ground contacts, a precise reference maneuver profile should be generated prior to targeting. Consequently, we present a technique, optimized via data-driven learning, for streamlining the maneuver profile connecting target-oriented positions. Marine biology Employing a bidirectional long short-term memory deep neural network, we modeled the quaternion profiles of low Earth orbit satellites. The model's function was to anticipate the maneuvers between target-pointing attitudes. After the attitude profile was predicted, the calculations for the time and angular acceleration profiles ensued. The optimal maneuver reference profile resulted from the application of Bayesian-based optimization. An investigation into the efficacy of the suggested technique involved scrutinizing the results of maneuvers ranging from 2 to 68.
This paper details a novel method for the continuous operation of a transverse spin-exchange optically pumped NMR gyroscope, achieved through modulating both the applied bias field and the optical pumping process. We report the simultaneous, continuous excitation of 131Xe and 129Xe using a hybrid modulation method, coupled with real-time demodulation of the Xe precession signal via a specialized least-squares fitting algorithm. Measurements of rotational speed are provided by this device, exhibiting a common field suppression factor of 1400, an angle random walk of 21 Hz/Hz, and a bias instability of 480 nHz after 1000 seconds.
For complete coverage path planning, the mobile robot must navigate through every attainable point documented within the environmental map. Considering the issues of suboptimal local paths and inadequate path coverage in complete coverage path planning using conventional biologically inspired neural networks, a Q-learning-based path planning algorithm for complete coverage is developed. The reinforcement learning methodology used in the proposed algorithm introduces the global environmental information. Sodium Bicarbonate Furthermore, the Q-learning approach is employed for path planning at points where accessible path points fluctuate, thereby enhancing the original algorithm's path planning strategy in the vicinity of such obstacles. Simulation results indicate the algorithm's capability to autonomously generate a well-structured pathway within the environmental map, achieving full coverage with a minimal rate of path repetition.
The escalating assault on traffic signals across the globe emphasizes the crucial role of intrusion detection. Traffic signal Intrusion Detection Systems (IDSs), utilizing data from connected cars and image processing, are restricted to detecting intrusions engineered by vehicles utilizing deceptive tactics. Nevertheless, these strategies are inadequate for identifying incursions launched against sensors located on roadways, traffic control units, and signal systems. In this paper, we propose an IDS that identifies anomalies in flow rate, phase time, and vehicle speed. This constitutes a substantial extension of our prior work, incorporating supplementary traffic data and statistical analysis. Based on the Dempster-Shafer decision theory, our system's theoretical model considered the current traffic parameters and their historical norms. We further utilized Shannon's entropy to evaluate the degree of uncertainty embedded in the observations. To validate our findings, a simulation model was designed using the SUMO traffic simulator and was populated with data from many real-world scenarios, gathered by the Victorian Transport Authority, Australia. Attacks such as jamming, Sybil, and false data injection were factored into the generation of scenarios for abnormal traffic conditions. The findings demonstrate that our proposed system achieves a remarkable 793% detection accuracy, minimizing false alarms.
Sound source characteristics, such as presence, location, type, and trajectory, are readily attainable through acoustic energy mapping. For this intention, different beamforming-oriented procedures can be employed. However, the timing discrepancies of the signals' arrival at every recording node (or microphone) dictate the necessity for synchronized multi-channel recordings. The practical application of a Wireless Acoustic Sensor Network (WASN) is evident when used to map the acoustic energy of an acoustic environment. Nonetheless, a characteristic concern relates to the inconsistent synchronization between the recordings from every node. This paper seeks to characterize the impact of today's popular synchronization methods, used within the context of WASN, to gather precise data for constructing acoustic energy maps. In the synchronization protocol evaluation, Network Time Protocol (NTP) and Precision Time Protocol (PTP) were compared. Three different audio capture methods were suggested for the WASN acoustic signal acquisition, two of which focused on local data storage and one on transmission through a local wireless network. A Wireless Acoustic Sensor Network (WASN), designed for practical evaluation, was built using Raspberry Pi 4B+ nodes, each incorporating a single MEMS microphone. The experimental data definitively underscores the robustness of the PTP synchronization protocol, coupled with local audio recording, as the most reliable approach.
In light of the unavoidable risks stemming from operator fatigue in present ship safety braking methods' dependence on ship operators' driving, this study endeavors to reduce the negative impact on navigation safety. Using a functional and technical approach, a human-ship-environment monitoring system was established in this initial study. The investigation into a ship braking model is pivotal, and this model incorporates electroencephalography (EEG) for monitoring brain fatigue, thus decreasing braking-related safety risks during navigation. Afterwards, the Stroop task experiment was adopted to evoke fatigue responses in drivers. This research project utilized principal component analysis (PCA) to streamline data dimensionality across multiple channels of the data acquisition device, isolating centroid frequency (CF) and power spectral entropy (PSE) features from channels 7 and 10. Moreover, a correlation analysis was carried out to examine the connection between these factors and the Fatigue Severity Scale (FSS), a five-point rating scale for assessing the degree of fatigue experienced by the subjects. By employing ridge regression and focusing on the three features exhibiting the highest correlation, this study created a model for determining driver fatigue levels. The proposed human-ship-environment monitoring system, coupled with a fatigue prediction model and ship braking model, facilitates a safer and more controllable ship braking process in this study. Predictive and real-time monitoring of driver fatigue allows for timely interventions ensuring navigation safety and driver well-being.
The current development of artificial intelligence (AI) and information and communication technology is causing a transformation in ground, air, and sea vehicles from human-controlled to unmanned, operating without human involvement. Unmanned surface and underwater vehicles, collectively known as unmanned marine vehicles (UMVs), can complete maritime tasks that are presently unachievable by manned vessels, decreasing personnel risk, enhancing power requirements for military missions, and yielding substantial economic benefits. The purpose of this review is to uncover historical and current trends in UMV development, and to present forward-looking perspectives on future UMV developments. Unmanned maritime vehicles (UMVs) are scrutinized in the review, showcasing their potential benefits including completing maritime tasks which are currently beyond the capabilities of crewed vessels, diminishing the risk linked to human presence, and amplifying capabilities for military assignments and economic advancement. Unmanned Vehicles (UVs) utilized in the air and on the ground have witnessed faster advancement compared to Unmanned Mobile Vehicles (UMVs) in view of the challenging operational environments for UMVs. The challenges encountered in the development of unmanned mobile vehicles, particularly within challenging environments, are highlighted in this review. Continued advancements in communication and networking, navigation and sound exploration, and multi-vehicle mission planning technologies are crucial for enhancing unmanned vehicle collaboration and intelligence.