In the field of mobile robotics, point clouds are widely used as an environment representation merged into a map. However, point cloud registration can be challenging when rapid movement occurs. In indoor environments...
详细信息
ISBN:
(纸本)9798350362350;9798350362343
In the field of mobile robotics, point clouds are widely used as an environment representation merged into a map. However, point cloud registration can be challenging when rapid movement occurs. In indoor environments, planelike features such as walls, floors, and ceilings are dominant and can be used to aid cloud merging. these features need to be extracted from point clouds. In this paper, we propose an open-source Hough transform implementation in Python. We have adapted this method, conventionally used for detecting shapes in images, to be applied to 3D point clouds and to detect planar features. To illustrate the point-cloud Hough transform algorithm, we generated synthetic point clouds. Additionally, we collected LiDAR measurements in the corridor. Processing the experimental data proved the algorithms ability to handle noise and measurement uncertainties. the study included an evaluation of the Hough transform by making a comparison to the ICP method. the Hough Transform provided more accurate results of corridor width measured on a merged point cloud by 0.06 m.
there is method for real time frequency analysis of the electromyogram is described in article. this method is proposed to use in control systems of advanced robotics for medical and healthcare applications. Method co...
详细信息
ISBN:
(纸本)9781467355087;9781467355063
there is method for real time frequency analysis of the electromyogram is described in article. this method is proposed to use in control systems of advanced robotics for medical and healthcare applications. Method consists of real-time frequency estimation with preliminary using of band-pass filter. the method is easy to program and does not require any additional resources. the article gives an example of using this method in classification of movements problem which is actual for robotics and human-machine interfaces.
In this paper a method is introduced that combines Inertial Measurement Unit (IMU) readouts with low accuracy and temporarily unavailable velocity measurements (e.g., based on kinematics or GPS) to produce high accura...
详细信息
ISBN:
(纸本)9781479987016
In this paper a method is introduced that combines Inertial Measurement Unit (IMU) readouts with low accuracy and temporarily unavailable velocity measurements (e.g., based on kinematics or GPS) to produce high accuracy estimates of velocity and orientation with respect to gravity. the method is computationally cheap enough to be readily implementable in sensors. the main area of application of the introduced method is mobile robotics.
the issue of identifying and designing control systems of fractional order models involves the necessity of their approximation to integer order. the most popular techniques include: Oustaloup's method, Matsuda...
详细信息
ISBN:
(纸本)9798350362350;9798350362343
the issue of identifying and designing control systems of fractional order models involves the necessity of their approximation to integer order. the most popular techniques include: Oustaloup's method, Matsuda's method, CFE (Continued Fraction Expansion) method, M-SBL method and different more. It is also possible to use the Pade approximation. the article examines some properties of this method. the case of the fractional power term (complex variable transfer function raised to a fractional power) is the main focus since it serves as a basis for approximating more intricate fractional order systems. the article's innovation is in using the nonlinear least squares method to obtain a mathematical description of the behavior of the zeros and poles of the chosen component after approximation. Relying on the gathered statistics, it was found that a very good fit was produced. the approximation of the behaviour of zeros and poles also allowed for the presentation of a new, alternative notation for the Pade approximation.
We present a comparative case study of machine learning models, evaluating their efficiency in a practical task of multiclass classification of samples being submissions to a recruitment survey and assigning them scor...
详细信息
ISBN:
(纸本)9798350311075
We present a comparative case study of machine learning models, evaluating their efficiency in a practical task of multiclass classification of samples being submissions to a recruitment survey and assigning them scores denoting the match level for a given candidate to a given workgroup (committee) in the AGH Students' Council. this research is based on the Council's recruitment applications that carried candidates' responses to a set of 10 hypothetical Council member activity scenarios, where they were to choose one of four given solutions to the problems. the data was collected from a web quiz in 2020, validated on a voluntary insider control group's responses to these questions and finally the best-performing model was evaluated in practice in the 2021's recruitment inside an in-browser adventure minigame. this work provides insight into how models ranging from classical methods to deep learning perform in a very specific not yet well-explored in literature, practical non-linear problem that is dependent on individual features of the participants, withthe data volume being very limited due to a restricted population of candidates. this information may provide a starting point for applications of machine learning in decision support systems in recruitment processes.
the problem of PID type controller tuning has been addressed in this paper. In particular, a method of selection of PD settings based on the solution of linear-quadratic optimisation problem using the energy criterion...
详细信息
ISBN:
(纸本)9781728173818;9781728173801
the problem of PID type controller tuning has been addressed in this paper. In particular, a method of selection of PD settings based on the solution of linear-quadratic optimisation problem using the energy criterion has been investigated. thus, the possibility of transforming optimal settings of the linear-quadratic regulator into the settings of the controller in the classical control system has been given. the presented methodology has been used during synthesis of control system for a two-wheeled balancing robot. Finally, the performance of the proposed control system has been validated by simulation in Matlab/Simulink environment withthe use of a two-wheeled balancing robot model.
Optimal control problem for semilinear hyperbolic equations is considered on 3 -star metric graph. the distributed controls on the edges are considered. the tracking type cost for steady state equation is considered f...
详细信息
ISBN:
(纸本)9798350362350;9798350362343
Optimal control problem for semilinear hyperbolic equations is considered on 3 -star metric graph. the distributed controls on the edges are considered. the tracking type cost for steady state equation is considered for the control problem. Using the Turnpike Property for dynamic control problem, the control problem is solved for the steady state equations. the necessary and sufficient optimality conditions are established for static control problems. then the optimum design problem is considered for the topological perturbation of the graph. the small cycle of size epsilon -> 0 replaces the central node of 3-star graph. the distributed controls are introduced on the edges of the cycle. the topological derivative of the shape functional indicates the possibility of nucleation of the small cycle.
Deep Neural Networks, particularly Convolutional Neural Networks (ConvNets), have achieved incredible success in many vision tasks, but they usually require millions of parameters for good accuracy performance. With i...
详细信息
ISBN:
(数字)9781665490627
ISBN:
(纸本)9781665490627
Deep Neural Networks, particularly Convolutional Neural Networks (ConvNets), have achieved incredible success in many vision tasks, but they usually require millions of parameters for good accuracy performance. With increasing applications that use ConvNets, updating hundreds of networks for multiple tasks on an embedded device can be costly in terms of memory, bandwidth, and energy. Approaches to reduce this cost include model compression and parameter-efficient modelsthat adapt a subset of network layers for each new task. this work proposes a novel parameter-efficient kernel modulation (KM) method that adapts all parameters of a base network instead of a subset of layers. KM uses lightweight task-specialized kernel modulators that require only an additional 1.4% of the base network parameters. With multiple tasks, only the task-specialized KM weights are communicated and stored on the end-user device. We applied this method in training ConvNets for Transfer Learning and Meta-Learning scenarios. Our results show that KM delivers up to 9% higher accuracy compared to other parameter-efficient methods on the Transfer Learning benchmark.
In this paper new results on the application of the Iterative Learning Control (ILC) schemes applied to the Gantry robot are presented. the main result given refers to the situation where the reference signal pre-defi...
详细信息
ISBN:
(纸本)9798350362350;9798350362343
In this paper new results on the application of the Iterative Learning Control (ILC) schemes applied to the Gantry robot are presented. the main result given refers to the situation where the reference signal pre-defined for ILC can be extended by some quantity that genuinely does not appear in the model but is important for the practical application of such system. In case of a considered system - the Gantry robot - its original state space model output refers to the robot position that in turn defines the robot's movement trajectory. For many practically-relevant applications it is enough to define the reference signal for the ILC in terms of this trajectory. this is not always ideal, as for some tasks it is easier (at least in terms of wear of the robot parts) to include additional requirements for the reference signal. the results presented in this paper allow to extend it by an additional factor - which, in this case, is the robot velocity. In what follows the ILC scheme is designed and implemented for such a combined reference signal.
Model Predictive Control (MPC) is an advanced method of process control. Despite its usefulness, it is applied mostly for large industrial processes. In the paper, a model predictive algorithm for a glass forehearth i...
详细信息
ISBN:
(纸本)9781728173818;9781728173801
Model Predictive Control (MPC) is an advanced method of process control. Despite its usefulness, it is applied mostly for large industrial processes. In the paper, a model predictive algorithm for a glass forehearth is presented. the problem of molten glass temperature stabilisation under external disturbances is especially important during the glass conditioning, so the use of an adaptive predictive controller seems to be reasonable. the controller tuning utilizes linear models of the process, that can be obtained on-line. Modifications of the known continuous-time MPC approach are described. the most important difference is the original method of measurable disturbances compensation and its implementation in the algorithm. the developed controller was tested using the process model with distributed parameters (Partial Differential Equation). the experimental results are presented in the paper.
暂无评论