38139 research outputs found
Sort by
A Trajectory Planning and Tracking Method Based on Deep Hierarchical Reinforcement Learning
No full textTo improve the driving efficiency of unmanned vehicles in a complex urban traffic flow environment and the safety and passenger comfort of vehicles when changing lanes, we propose a hierarchical reinforcement learning (HRL)-based vehicle trajectory planning and tracking method. First, we present a hierarchical control framework for vehicle trajectory tracking that is based on deep reinforcement learning (DRL) and model predictive control (MPC). We design an upper-level decision model based on the trust region policy optimization algorithm integrated with long short-term memory to obtain more accurate strategies. Second, to improve stability and passenger comfort, we constructed a lower controller that combines the Bezier curve fitting method and an MPC controller. Finally, the proposed method was simulated via the car learning to act (CARLA) simulator, which is based on an unreal engine. Random urban traffic-flow test scenarios were used to simulate a real urban road-traffic environment. The simulation results illustrate that the proposed method can complete the vehicle trajectory planning and tracking task well. Compared with the existing RL methods, our proposed method has the lowest collision rate of 1.5% and achieves an average speed improvement of 7.04%. Moreover, our proposed method has better comfort performance and lower fuel consumption during the driving process
12th International Rangeland Congress
No full textThe highly variable weather and climate of northern Australia can pose a significant threat to cattle and other livestock, with prolonged heat waves and sudden chill conditions known to increase mortality risk. For example, the compounding impact of high temperatures, high humidity and calm conditions led to significant cattle heat stress and dozens of animal deaths in southern Queensland in late January 2024. Conversely, the combination of flooding, low temperatures, and high winds associated with a tropical low caused thousands of cattle deaths in northern Queensland in February
Currently, the Australian Bureau of Meteorology issues national sheep graziers' alerts for potential risk of chill Test and exposure, however there are no such equivalent chill (or heat) warnings for cattle. A key objective of the Northern Australian Climate Program (NACP) is to develop prototype forecast products of thermal stress that can be utilised by livestock producers to help manage the risks posed by extreme weather and climate events.
In this research, we describe NACP's latest prototype forecast maps of the Heat Load Index (HLI) and Cattle Comfort Index (CCI), derived from the Bureau's numerical weather prediction system - ACCESS-G3. These forecasts display the predicted chill and heat conditions across Australia out to 7 days. We also assess how well the predictions of CCI performed for a extreme heat event in southern Queensland in January 2024
A systematic scoping review of patient-specific devices in operative management of scaphoid nonunion
Full text linkScaphoid nonunion is a complication that occurs in 5-10% of patients who sustain an acute scaphoid fracture. There is no superior method for managing scaphoid nonunion that has been established in the current literature. The use of patient specific devices has been explored as a way to improve surgical management of scaphoid nonunion. 3D printing has made the production of patient specific devices more accessible. CT and MRI images can be used to model devices such as anatomic models and surgical guides. In this review we aim to discuss the current evidence for the use of patient specific devices for managing chronic scaphoid nonunion and demonstrate their different uses. A scoping review of peer-reviewed literature, existing patent applications, and grey literature was performed. Although the current research has demonstrated some possible benefits such as reduction in surgical time and accuracy of restoration of anatomy, further investigation with comparative studies is required to make conclusions about the superiority of these techniques over standard freehand techniques
ALMA millimetre-wavelength imaging of HD 138965: new constraints on the debris dust composition and presence of planetary companions
Full text linkHD 138965 is a young A type star and member of the nearby young Argus association. This star is surrounded by a broad, bright debris disc with two temperature components that was spatially resolved at far-infrared wavelengths by Herschel. Here, we present Atacama Large Millimiter/Submillimeter Array (ALMA) millimetre-wavelength imaging of the cool outer belt. These reveal its radial extent to be 150 au with a width () of 49 au (), at a moderate inclination of 499. Due to the limited angular resolution, signal-to-noise and inclination we have no constraint on the disc's vertical scale height. We modelled the disc emission with both gravitational and radiation forces acting on the dust grains. As the inner belt has not been spatially resolved, we fixed its radius and width prior to modelling the outer belt. We find astronomical silicate is the best fit for the dust composition. However, we could not reject possible scenarios where there are at least 10 per cent water-ice inclusions. Combining the spatially resolved imaging by ALMA with non-detection at optical wavelengths by HST, we obtain a limit on the scattering albedo for the debris dust in the outer belt. Analysis of the outer belt's architecture in conjunction with simple stirring models places a mass limit of 2.3 0.4 on a companion interior to the belt (78 au), a factor of two improvement over constraints from high contrast imaging
Occupational Therapy Australia 31st National Conference and Exhibition 2025 (OTAUS 2025)
No full textIntroduction
Traditional teaching methods that focus primarily on theoretical concepts mean students are not prepared for real-world contexts. Thus, learning tools that bridge the gap between theory and practice are essential. Given geographical, economic and time constraints faced by regional and remote Universities, it is not feasible to have industry professionals regularly attend face-to-face learning modalities, , limiting available learning experiences for students. Hence, online or virtual learning opportunities can enhance students’ experiences by providing choice and accessibility to essential resources.
Objectives
This review aimed to explore the evidence for utilising a multimodal pedagogical approach and innovative teaching practices such as podcasts and vignettes to enhance the student learning experience.
Methods
A scoping review was conducted using the Arskey and O’Malley (2005) five-stage framework. Sixteen studies were identified as suitable and met the inclusion and exclusion criteria.
Results
Podcasts may assist in building a narrative, strengthening the learning experience, and enhancing engagement with students. In contrast to an audiovisual file such as a lecture, which delivers information at a single point in time from a single viewpoint; a podcast brings the listener on a dynamic journey with the hosts and can address current and relevant topics. At a tertiary level, podcasts are an effective learning tool and are valued by students for being educational, entertaining, and portable.
Conclusion
Students are more likely to engage in learning core or challenging concepts in their discipline when provided with student-centred learning tools. Students report improved engagement and satisfaction with multimodal pedagogy such as podcasts
Prenatal depression level prediction using ensemble based deep learning model
Full text linkBackground and objective:
Many people find that the emotional and mental strain of labor and delivery is greater than they anticipated. However, there are few reports on stress levels during pregnancy, and there is limited research into stress observation during delivery. Prenatal depression during the delivery has to be monitored continuously without disturbing the mothers during the childbirth.
Methods:
We explore the potential of employing EDA for Prenatal Depression prediction. The proposed model applies a novel method for motion artifacts followed by data labeling using PHQ-9 score values and LOOCV applied to train robustly. This culminated in the development of a novel EBDL model to accurately predict stress levels.
Results:
We subsequently applied the ensemble based deep learning model on a testing dataset and our method proved to be 93.87 percent accurate, proving its superiority over the standard supervised classification models. The accuracy of this approach applied to three benchmark datasets produced better results compared to all commonly applied machine learning models, including an Ensemble based Deep Learning model.
Conclusion:
The preliminary results are promising, and indicate a superior utility of EDA for monitoring stress levels in real-life scenarios. This approach should be applied to a clinical setting, it potentially could continuously monitor stress levels in pregnant women and provide real-time feedback of clinically important data for clinicians
Developments and future prospects of welding technology for carbon fiber thermoplastic composites
No full textCarbon fiber reinforced thermoplastic composites (TPCs) attracted significant attentions from the aerospace, transportation, and defense industries, due to their high specific stiffness and specific strength, outstanding thermal stability and good damage resistance, etc. As the demand of TPCs significantly increased for aerospace applications, the development of advanced joining technologies for TPC components becomes critical to ensure the structural integrity of aviation structures. This paper provides a comprehensive review of the historical development and recent advancements in welding technologies for TPCs, including ultrasonic welding, induction welding, resistance welding, and laser welding. Special emphasis is placed on ultrasonic welding due to its growing prominence in the field. The characteristics of various types of welding technologies for TPCs have been systematically discussed. Simultaneously, the strengths of the TPC joints manufactured by different welding technologies have been summarized and compared. The future development trend and research focuses for the welding technologies of TPC components are also proposed
Efficient production of syngas and lactic acid via CrB MBene/Cd0.8Zn0.2S Schottky heterojunction photocatalysis
Full text linkThe photocatalytic conversion of biomass utilizing two-dimensional (2D) MBene as co-catalyst represents an exemplary approach for the sustainable production of syngas and lactic acid, capitalizing on the renewable resources of biomass carbon and solar energy. In this study, CrB MBene was synthesized via in-situ exfoliation in an HCl aqueous solution and subsequently integrated with Cd0.8Zn0.2S through an in-situ growth strategy, aiming to facilitate the photocatalytic co-production of lactic acid and syngas. Notably, the CZS/CrB-10 variant demonstrated a remarkable 12.3-fold enhancement in the hydrogen evolution rate (1746.6 μmol g−1 h−1) compared to standalone Cd0.8Zn0.2S, highlighting the exceptional co-catalytic efficiency and air stability of CrB MBene. Under optimal experimental conditions, the lactic acid yield and CO evolution rate attained 75.4 % and 917.8 μmol g−1 h−1, respectively. The incorporation of CrB MBene as a co-catalyst was found to significantly improve sunlight utilization, enhance the separation and transfer of photogenerated charge carriers, and mitigate the recombination of electrons and holes. This research underscores the potential of CrB MBene as a promising and cost-effective co-catalyst, with implications for advancing artificial photosynthetic systems and the integration of photocatalytic biomass refining with water splitting technologies
Superhydrophobic fire-extinguishing polyurethane foam for solar-assisted high-efficiency recovery of viscous crude oil spill
No full textFrequent offshore crude oil spill accidents pose a significant threat to marine ecosystems and coastal communities. Due to the high viscosity and poor fluidity of crude oil, there is an urgent need for polyurethane foam with excellent photothermal properties and oil–water separation capabilities to facilitate crude oil absorption. However, the flammability of PU foam greatly restricts its application when faced with fire scenarios in offshore petrochemical spill incidents. To solve these challenges, a flame-retardant superhydrophobic-superoleophilic polyurethane foam (PDMS@PLP@MXene@PU) is designed by assembling dual photothermal layers and flame retardant onto the foam structure via electrostatic attraction and hydrogen bonding. The results show that PDMS@PLP@MXene@PU exhibits superhydrophobic properties (water contact angle = 162.4°) and crude oil absorption capacity (64.2 g/g). The compressive strength of the PDMS@PLP@MXene is enhanced by 83.9 %. PDMS@PLP@MXene@PU exhibits good photothermal effect and thermal conductivity, which can rapidly rise to 80.0 °C under 1 kW/m2 solar irradiation with a maximum oil absorption rate of ∼ 98 %. PDMS@PLP@MXene@PU can self-extinguish a flame with 52.7 % and 71.4 % reductions in peak heat release rate and total smoke production, respectively. This work offers a facile strategy for creating high-performance polyurethane foam to address crude oil spills
From Tweets to Threats: A Survey of Cybersecurity Threat Detection Challenges, AI-Based Solutions and Potential Opportunities in X
No full textThe pervasive use of social media platforms, such as X (formerly Twitter), has become a part of our daily lives, simultaneously increasing the threat of cyber attacks. To address this risk, numerous studies have explored methods to detect and predict cyber attacks by analyzing X data. This study specifically examines the application of AI techniques for predicting potential cyber threats on X. DeepNN consistently outperforms competing methods in terms of overall and average figure of merit. While character-level feature extraction methods are abundant, we contend that a semantic focus is more beneficial for this stage of the process. The findings indicate that current studies often lack comprehensive evaluations of critical aspects such as prediction scope, types of cybersecurity threats, feature extraction techniques, algorithm complexity, information summarization levels, scalability over time, and performance measurements. This review primarily focuses on identifying AI methods used to detect cyber threats on X and investigates existing gaps and trends in this area. Notably, over the past few years, limited review articles have been published on detecting cyber threats on X, especially those concentrating on recent journal articles rather than conference papers