Modeling 3D objects in domains like computer Aided Design (CAD) is time-consuming and comes with a steep learning curve needed to master the design process as well as tool complexities. In order to simplify the modeli...
详细信息
ISBN:
(纸本)9789897584886
Modeling 3D objects in domains like computer Aided Design (CAD) is time-consuming and comes with a steep learning curve needed to master the design process as well as tool complexities. In order to simplify the modeling process, we designed and implemented a prototypical system that leverages the strengths of Virtual Reality (VR) hand gesture recognition in combination with the expressiveness of a voice-based interface for the task of 3D modeling. Furthermore, we use the Constructive Solid Geometry (CSG) tree representation for 3D models within the VR environment to let the user manipulate objects from the ground up, giving an intuitive understanding of how the underlying basic shapes connect. The system uses standard mid-air 3D object manipulation techniques and adds a set of voice commands to help mitigate the deficiencies of current hand gesture recognition techniques. A user study was conducted to evaluate the proposed prototype. The combination of our hybrid input paradigm shows to be a promising step towards easier to use CAD modeling.
A reinforcement learning agent for optimal green-house management through Proximal Policy Optimisation is developed and evaluated in this work. With variable elements like lighting, irrigation, humidity, and temperatu...
A reinforcement learning agent for optimal green-house management through Proximal Policy Optimisation is developed and evaluated in this work. With variable elements like lighting, irrigation, humidity, and temperature, a virtual greenhouse is produced. By interacting with the surroundings, the agent modifies the controls and gets input on things like crop health and growth rate. For policy and value estimation, the neural network design consists of distinct actor and critic networks as well as shared feature layers. Compared to fixed or random policies, the trained agent shows the ability to regulate conditions for increased crop yield and sustainability. As the number of trials increased, the RL agent's average return per trial increased steadily. This suggests that the RL agent has become more adept at controlling the greenhouse environment in subsequent experiments. This demonstrates how well the RL agent can manage crop health and resource use in the greenhouse climate control system.
The control of indoor temperature has significant importance to maintain excellent thermal comfort and energy saving. At present, indoor temperature is mainly controlled by Air-Handling Unit (AHU), so a proper indoor ...
详细信息
With the development of fuel cell technology, how to effectively reduce hydrogen consumption of fuel cell vehicles has become a technical focus and difficulty. This paper proposed an adaptive energy management algorit...
详细信息
With the development of fuel cell technology, how to effectively reduce hydrogen consumption of fuel cell vehicles has become a technical focus and difficulty. This paper proposed an adaptive energy management algorithm which was suitable for fuel cell vehicles. First, the ECMS algorithm considering the durability of the fuel cell was established. In order to ensure the good performance of the ECMS algorithm, an equivalent factor optimization model was established, which adjusted the equivalent factor according to the changes of battery SOC and future velocity. Considering that in the actual driving environment, it was impossible to accurately obtain the upcoming driving velocity in the future. Therefore, it was hard to quickly adjust and select the equivalent factor. This article established a neural network velocity prediction model, and the predicted velocity were provided to the equivalent factor optimization model, and the adaptive energy management algorithm was accomplished. Simulation results showed that the adaptive algorithm reduced hydrogen consumption by 2.6% compared to the basic ECMS algorithm, while battery SOC was in a suitable wording window.
Deep getting to know techniques are turning into extra and extra famous in the area of goal monitoring due to their effective function extraction capabilities, amongst which the twin network-based goal monitoring algo...
Deep getting to know techniques are turning into extra and extra famous in the area of goal monitoring due to their effective function extraction capabilities, amongst which the twin network-based goal monitoring algorithm with excessive walking body price and robust monitoring fault tolerance has attracted the interest of many scholars. Aiming at the drawbacks that twin networks use an offline method to recognize end-to-end mannequin training, and the educated mannequin can't trade the community parameters and the goal template at some point of the monitoring process, this paper proposes an increased twin community goal monitoring algorithm primarily based on a hybrid interest mechanism, which is first off skilled offline on the education set of the ILSVRC2015 dataset, and then on the coaching units of ILSVRC2015 and The algorithm is examined and evaluated on a take a look at set on the plane class of the VOT2018 dataset, and the experimental outcomes exhibit that the success fee of the algorithm proposed in this paper is multiplied from 61.2% to 68.4% and the accuracy from 77.1% to 82.3% of the SiamFC base algorithm, which proves that the algorithm has properly monitoring performance.
In this paper, we focus on recognizing 3D shapes from arbitrary views, i.e., arbitrary numbers and positions of viewpoints. It is a challenging and realistic setting for view-based 3D shape recognition. We propose a c...
ISBN:
(纸本)9781665428125
In this paper, we focus on recognizing 3D shapes from arbitrary views, i.e., arbitrary numbers and positions of viewpoints. It is a challenging and realistic setting for view-based 3D shape recognition. We propose a canonical view representation to tackle this challenge. We first transform the original features of arbitrary views to a fixed number of view features, dubbed canonical view representation, by aligning the arbitrary view features to a set of learnable reference view features using optimal transport. In this way, each 3D shape with arbitrary views is represented by a fixed number of canonical view features, which are further aggregated to generate a rich and robust 3D shape representation for shape recognition. We also propose a canonical view feature separation constraint to enforce that the view features in canonical view representation can be embedded into scattered points in a Euclidean space. Experiments on the ModelNet40, ScanObjectNN, and RGBD datasets show that our method achieves competitive results under the fixed viewpoint settings, and significantly outperforms the applicable methods under the arbitrary view setting.
In this current era, usage of the internet is increasing drastically. Digitization has led to high network traffic which makes the overall management of network highly complex and expensive as traditional networks are...
详细信息
This project proposes an intelligent aquatic lifesaving robot based on computervision and machine learning technology. The robot realizes high-precision positioning through multi-satellite positioning function, and i...
详细信息
ISBN:
(纸本)9798400708299
This project proposes an intelligent aquatic lifesaving robot based on computervision and machine learning technology. The robot realizes high-precision positioning through multi-satellite positioning function, and is equipped with posture sensing function and computervision function. It uses Convolutional neural network and Convolutional Pose Machine algorithm to realize human posture recognition. At the same time, the robot has the function of self-balance adjustment and wireless charging, which improves the stability and sustainability of the robot. The experimental results show that the effect of action recognition using 3D convolutional neural network is better than the traditional k-nearest neighbor classifier and support vector machine classifier. In the Weizmann action database, the recognition rate of 3D convolutional neural network reaches 96.3%, which is significantly higher than 90.6% of K-nearest classifier and 93.1% of support vector machine classifier. In the KTH action database, the recognition rate of 3D convolutional neural network is 90.3%, which is higher than 86.7% of K-nearest classifier and 89% of support vector machine classifier.
The problem of a control system design is discussed to provide an angular position platform stabilization. The platform is used to install various navigation equipment and communication devices. Stabilization of the p...
详细信息
ISBN:
(数字)9798331516321
ISBN:
(纸本)9798331516338
The problem of a control system design is discussed to provide an angular position platform stabilization. The platform is used to install various navigation equipment and communication devices. Stabilization of the platform's angular position eliminates the influence of vibrations of the platform's moving base on the operation of this equipment. The platform is equipped with an electrohydraulic drive, which ensures the horizontal position of the platform under conditions of vibrations of the vessel deck. The control system is designed in such a way that fast and slow processes are created in a closedloop system. The discussed approach to control system design allows to provide stability and robustness for position of the platform under vibrations of the vessel deck. The effect of the small parameters of the electrohydraulic drive on control system stability is treated. The selection of the controller parameters is carried out by additionally taking into account the sensitivity function, which reflects the influence of wave disturbances on the angular position of the platform.
This paper recommends a load frequency controller (LFC) for integrating non-conventional energy resources into marine vessel’s power system. The LFC, based on a conventional proportional integral derivative (PID) con...
This paper recommends a load frequency controller (LFC) for integrating non-conventional energy resources into marine vessel’s power system. The LFC, based on a conventional proportional integral derivative (PID) controller optimized using particle swarm optimization (PSO), addresses frequency deviations caused by non-conventional energy sources and non-linear dynamic loads in isolated hybrid marine microgrid. The controller's design considers transfer functions of microgrid components, while its performance is evaluated using integral time absolute error (ITAE) and integral time square error (ITSE) indices. Validation is conducted through real-time simulations using the DS 1104 R&D controller board. The operational results demonstrate the superior performance of the tuned LFC in regulating frequency deviations, with improved rise time, fall time, slew rate, and overshoot compared to the non-optimized controller. This research presents a resilient and efficient LFC design that effectively addresses frequency deviations in marine microgrids through the integration of non-conventional energy resources. The findings may contribute in advancing the sustainable operation of marine vessels, enabling energy source diversification and a reduction in greenhouse gas emissions.
暂无评论