human locomotion and activity recognition systems form a critical part in a robot's ability to safely and effectively operate in a environment populated with human end users. Previous work in this area relies upon...
详细信息
ISBN:
(纸本)9781509037636
human locomotion and activity recognition systems form a critical part in a robot's ability to safely and effectively operate in a environment populated with human end users. Previous work in this area relies upon strong assumptions about the labels in the training data;e.g. that are noise-free and that they exist at all. Our approach does not predefine the relevant behaviours or their number, as both are learned directly from observations, similar to real-world human-robot interactions, where labels are neither available. Instead we introduce models that make no assumptions about the state space, by presenting a fully unsupervised nonparametric Bayesian recognition approach, in which we leverage recent advances in state space modelling with automatic inference using probabilistic programming. We demonstrate the utility of full model optimisation using Bayesian optimisation and validate our approach on several challenging problems, using different feature modalities.
Stiffness modulus is a fundamental parameter used in the modelling of the viscoelastic behaviour of bituminous mixtures. On the basis of the master curve in the linear viscoelasticity range, the mechanical properties ...
Stiffness modulus is a fundamental parameter used in the modelling of the viscoelastic behaviour of bituminous mixtures. On the basis of the master curve in the linear viscoelasticity range, the mechanical properties of asphalt concrete at different loading times and temperatures can be predicted. This paper discusses the construction of master curves under rheological mathematical models i.e. the sigmoidal function model (MEPDG), the fractional model, and Bahia and co-workers' model in comparison to the results from mechanistic rheological models i.e. the generalized Huet-Sayegh model, the generalized Maxwell model and the Burgers model. For the purposes of this analysis, the reference asphalt concrete mix (denoted as AC16W) intended for the binder coarse layer and for traffic category KR3 (5×105
Mental models are pen pictures that are used to represent complex systems and their aspects including views of different system stakeholders. They are used extensively in complex system modelling and play a critical r...
详细信息
ISBN:
(纸本)9781509006212
Mental models are pen pictures that are used to represent complex systems and their aspects including views of different system stakeholders. They are used extensively in complex system modelling and play a critical role in model development. In this paper we investigate an echo state network (ESN) encoded with limited number of reservoir nodes to automatically learn mental models through the use of system observations. Three different evolutionary algorithms, genetic algorithm (GA), differential evolution (DE) and particle swarm optimisation (PSO) are used to optimise the weights and design parameters of ESN in order to learn models closer to manually developed models. Such an approach can be useful in reducing the modelling effort required by human modellers as well as the subjective bias in model development. The empirical analysis using two case studies shows that the ESN encoded with restricted number of nodes and optimised by different evolutionary algorithms guided with a combined fitness function that takes both output error and reservoir connectivity into account is able to learn similar structure as original mental models as well as is able to generate the correct output behaviour.
To address the impact of solar array (SA) anomalies and vibrations on performance of precision space-based operations, it is important to complete its accurate jitter analysis. This work provides mathematical modellin...
To address the impact of solar array (SA) anomalies and vibrations on performance of precision space-based operations, it is important to complete its accurate jitter analysis. This work provides mathematicalmodelling scheme to approximate kinematics and coupled micro disturbance dynamics of rigid load supported and operated by solar array drive assembly (SADA). SADA employed in analysis provides a step wave excitation torque to activate the system. Analytical investigations into kinematics is accomplished by using generalized linear and Euler angle coordinates, applying multi-body dynamics concepts and transformations principles. Theoretical model is extended, to develop equations of motion (EoM), through energy method (Lagrange equation). The main emphasis is to research coupled frequency response by determining energies dissipated and observing dynamic behaviour of internal vibratory systems of SADA. The disturbance model captures discrete active harmonics of SADA, natural modes and vibration amplifications caused by interactions between active harmonics and structural modes of mechanical assembly. The proposed methodology can help to predict true micro disturbance nature of SADA operating rigid load. Moreover, performance outputs may be compared against actual mission requirements to assess precise spacecraft controller design to meet next space generation stringent accuracy goals.
humanbehaviour has significant impact on the performance of any project. Software projects are no exception. This paper presents a research plan to develop combined agent-based and system dynamics models to simulate ...
详细信息
ISBN:
(纸本)9781450337960
humanbehaviour has significant impact on the performance of any project. Software projects are no exception. This paper presents a research plan to develop combined agent-based and system dynamics models to simulate software development process focusing on the behaviour of team members in a software project environment. The outcome of such research work can be used to explore the relationships between various factors, such as workload, stress, attitude, management approach, organisation process, as well as individual and team performance. It can also be used as a tool to help managers of software organisations to understand the importance of human aspects, and make better human resource management decisions for improved project performance.
Recently it has been noted that user behaviour can have a large impact on the final energy consumption in buildings. Through the combination of mathematicalmodelling and data from wireless ambient sensors, we can mod...
详细信息
ISBN:
(纸本)9781479985470
Recently it has been noted that user behaviour can have a large impact on the final energy consumption in buildings. Through the combination of mathematicalmodelling and data from wireless ambient sensors, we can model humanbehaviour patterns and use the information to regulate building management systems (BMS) in order to achieve the best trade-off between user comfort and energy efficiency. In this work, we have modelled user occupancy and activity patterns using Machine Learning approaches. We have applied non-linear multiclass Support Vector Machines (SVMs) to deal with the complex nature of the data collected from various sensors to accurately identify user occupancy and activities of daily living (ADL) patterns. To validate our results, we also used other methodologies (i.e. Hidden-Markov Model and k-Nearest Neighbours). The experimental results show that our proposed approach outperforms the other methods for the scenarios evaluated.
Rider models are employed to gain insight into bicycle rider steering behaviour and to improve characteristic properties of bicycles. In this paper, stability properties as well as basic dynamic characteristics of the...
详细信息
Rider models are employed to gain insight into bicycle rider steering behaviour and to improve characteristic properties of bicycles. In this paper, stability properties as well as basic dynamic characteristics of the passive (uncontrolled) bicycle rider system and consequences on the rider control modelling are addressed. In particular, the unstable motion of the system at low velocities and bandwidth limitations caused by non-minimum phase dynamics are emphasized. To analyse the effectiveness of the steering torque and the lean torque as possible rider's inputs to control the dynamics of the bicycle, a controllability analysis of the bicycle rider system has been performed. It turns out that lean torque input, in contrast to steering torque input, has marginal impact on the dynamics of the system Finally, a bicycle rider control model considering human rider properties is presented, and its capabilities are demonstrated by performing a curve entering manoeuvre. (C) 2015, IFAC (International Federation of Automatic Control) Hosing by Elsevier Ltd. All rights reserved.
Today's product development process is characterized by increasing complexity of products. Subsequently universities have to adapt and constantly improve the content of their courses to prepare the future graduate...
详细信息
ISBN:
(纸本)9781904670742
Today's product development process is characterized by increasing complexity of products. Subsequently universities have to adapt and constantly improve the content of their courses to prepare the future graduates for the free market economy. An approach at Technische Universitat Darmstadt to expand the scope of CAD education, is the introduction of a new one-week tutorial on shape design with Siemens NX. This paper will describe the teaching concept and its implementation. The concurrently submissions of the examinations with content from realistic industrial tasks play an important role during the course.
Weather conditions have an important influence on failure rates of Power Distribution Systems. An event with strong winds gust and a large number atmospheric discharge can increase the outages' quantity in a hard ...
详细信息
ISBN:
(纸本)9781509037933
Weather conditions have an important influence on failure rates of Power Distribution Systems. An event with strong winds gust and a large number atmospheric discharge can increase the outages' quantity in a hard to predictable form and, consequently, impact in the System Average Interruption Duration Index (SAIDI). A modelling of such weather impact in failure rate is useful for a better understand of it, allowing utilities to take actions to understatement it or, even, mitigate its consequences in real time. Based on it, this paper analyses the response of Negative Binomial Regression in comparison with the Poisson Regression, using outages data (quantity) from a Brazilian distribution utility and meteorological information as wind gust speed and number of thunders, to estimate the number of expected outages and, consequently, the failure rate per kilometer. The focus is to provide a model that emulates the behaviour of system components and parts during a climatic event of any intensity and has more resilient against the outliers normally presented in this type of data set. The results indicate a significant influence of the wind gust speed, and, based on goodness-of-fit index as Akaike Information Criterion and the analysis of the outliers, the Negative Binomial Regression as the best response. The estimated function can be useful to improve the utilities' response, helping them to improve their contingency plan.
The current design of steel-concrete composite floors might be susceptible to resonance phenomenon, causing undesirable vibrations in the frequency range that is the most noticeable to humans, i.e. 4Hz to 8Hz. In addi...
详细信息
暂无评论