scholarly journals The Model and Training Algorithm of Compact Drone Autonomous Visual Navigation System

Data ◽  
2018 ◽  
Vol 4 (1) ◽  
pp. 4 ◽  
Author(s):  
Viacheslav Moskalenko ◽  
Alona Moskalenko ◽  
Artem Korobov ◽  
Viktor Semashko
Keyword(s):  
Autonomous Navigation ◽  
Search Algorithm ◽  
Visual Navigation ◽  
Frame Rate ◽  
Raspberry Pi ◽  
Navigation Systems ◽  
Training Algorithms ◽  
Translation Error ◽  
Deep Model ◽  
And Training

Trainable visual navigation systems based on deep learning demonstrate potential for robustness of onboard camera parameters and challenging environment. However, a deep model requires substantial computational resources and large labelled training sets for successful training. Implementation of the autonomous navigation and training-based fast adaptation to the new environment for a compact drone is a complicated task. The article describes an original model and training algorithms adapted to the limited volume of labelled training set and constrained computational resource. This model consists of a convolutional neural network for visual feature extraction, extreme-learning machine for estimating the position displacement and boosted information-extreme classifier for obstacle prediction. To perform unsupervised training of the convolution filters with a growing sparse-coding neural gas algorithm, supervised learning algorithms to construct the decision rules with simulated annealing search algorithm used for finetuning are proposed. The use of complex criterion for parameter optimization of the feature extractor model is considered. The resulting approach performs better trajectory reconstruction than the well-known ORB-SLAM. In particular, for sequence 7 from the KITTI dataset, the translation error is reduced by nearly 65.6% under the frame rate 10 frame per second. Besides, testing on the independent TUM sequence shot outdoors produces a translation error not exceeding 6% and a rotation error not exceeding 3.68 degrees per 100 m. Testing was carried out on the Raspberry Pi 3+ single-board computer.

scholarly journals Autonomous navigation system with small four-wheel drive platform

2020 ◽  
Vol 166 ◽  
pp. 05004
Author(s):  
Martin Bogdanovskyi ◽  
Andrii Tkachuk ◽  
Oleksandr Dobrzhanskyi ◽  
Anna Humeniuk
Keyword(s):  
Autonomous Navigation ◽  
Hybrid Control ◽  
Vision System ◽  
Movement Control ◽  
Image Sensor ◽  
Obstacle Detection ◽  
Raspberry Pi ◽  
Navigation Systems ◽  
Wheel Drive ◽  
Four Wheel Drive

The task of achieving greater flexibility and maneuverability of small transport and service units’ motion in modern factories by developing small autonomous navigation systems plays crucial role in complex automation of transport logistics nowadays. To solve navigation task, it was proposed the following approach, where as a means of assessing the environment was used computer vision system based on 5-megapixel CMOS image sensor and for the front obstacle detection was used auxiliary ultrasonic sensor as a limit switch. Authors solved the problem of yawing using artificial marking approach as along two-colored leading lines. For maneuverability increase during the turn was used speed movement control based on lines perspective. The basic design and technical characteristics of the four-wheel drive platform and the algorithm of the Raspberry PI 3/Arduino Nano hybrid control system are presented. Experimental results proved the viability of the proposed approach.

Geophysical Support of Magnetometer Autonomous Navigation Systems

2020 ◽  
Author(s):  
Vladimir T. Minligareev ◽  
Elena N. Khotenko ◽  
Vadim V. Tregubov ◽  
Tatyana V. Sazonova ◽  
Vaclav L. Kravchenok
Keyword(s):  
Autonomous Navigation ◽  
Navigation Systems

scholarly journals VAGADRONE: Intelligent and Fully Automatic Drone Based on Raspberry Pi and Android

2021 ◽  
Vol 11 (7) ◽  
pp. 3153
Author(s):  
Saifeddine Benhadhria ◽  
Mohamed Mansouri ◽  
Ameni Benkhlifa ◽  
Imed Gharbi ◽  
Nadhem Jlili
Keyword(s):  
Remote Control ◽  
Autonomous Navigation ◽  
Object Identification ◽  
External Control ◽  
Raspberry Pi ◽  
Optimal Trajectories ◽  
Live Streaming ◽  
Wide Range ◽  
Fully Automatic ◽  
Direct Use

Multirotor drones are widely used currently in several areas of life. Their suitable size and the tasks that they can perform are their main advantages. However, to the best of our knowledge, they must be controlled via remote control to fly from one point to another, and they can only be used for a specific mission (tracking, searching, computing, and so on). In this paper, we intend to present an autonomous UAV based on Raspberry Pi and Android. Android offers a wide range of applications for direct use by the UAV depending on the context of the assigned mission. The applications cover a large number of areas such as object identification, facial recognition, and counting objects such as panels, people, and so on. In addition, the proposed UAV calculates optimal trajectories, provides autonomous navigation without external control, detects obstacles, and ensures live streaming during the mission. Experiments are carried out to test the above-mentioned criteria.

scholarly journals Air Learning: a deep reinforcement learning gym for autonomous aerial robot visual navigation

2021 ◽  
Author(s):  
Srivatsan Krishnan ◽  
Behzad Boroujerdian ◽  
William Fu ◽  
Aleksandra Faust ◽  
Vijay Janapa Reddi
Keyword(s):  
Reinforcement Learning ◽  
Embedded System ◽  
Broad Class ◽  
Visual Navigation ◽  
Raspberry Pi ◽  
Latency Distribution ◽  
Hardware In The Loop ◽  
Resource Constrained ◽  
Aerial Robot ◽  
Policy Optimization

AbstractWe introduce Air Learning, an open-source simulator, and a gym environment for deep reinforcement learning research on resource-constrained aerial robots. Equipped with domain randomization, Air Learning exposes a UAV agent to a diverse set of challenging scenarios. We seed the toolset with point-to-point obstacle avoidance tasks in three different environments and Deep Q Networks (DQN) and Proximal Policy Optimization (PPO) trainers. Air Learning assesses the policies’ performance under various quality-of-flight (QoF) metrics, such as the energy consumed, endurance, and the average trajectory length, on resource-constrained embedded platforms like a Raspberry Pi. We find that the trajectories on an embedded Ras-Pi are vastly different from those predicted on a high-end desktop system, resulting in up to $$40\%$$ 40 % longer trajectories in one of the environments. To understand the source of such discrepancies, we use Air Learning to artificially degrade high-end desktop performance to mimic what happens on a low-end embedded system. We then propose a mitigation technique that uses the hardware-in-the-loop to determine the latency distribution of running the policy on the target platform (onboard compute on aerial robot). A randomly sampled latency from the latency distribution is then added as an artificial delay within the training loop. Training the policy with artificial delays allows us to minimize the hardware gap (discrepancy in the flight time metric reduced from 37.73% to 0.5%). Thus, Air Learning with hardware-in-the-loop characterizes those differences and exposes how the onboard compute’s choice affects the aerial robot’s performance. We also conduct reliability studies to assess the effect of sensor failures on the learned policies. All put together, Air Learning enables a broad class of deep RL research on UAVs. The source code is available at: https://github.com/harvard-edge/AirLearning.

scholarly journals A Dynamic Calibration Method of Installation Misalignment Angles between Two Inertial Navigation Systems

Sensors ◽  
2018 ◽  
Vol 18 (9) ◽  
pp. 2947
Author(s):  
Ming Hua ◽  
Kui Li ◽  
Yanhong Lv ◽  
Qi Wu
Keyword(s):  
Autonomous Navigation ◽  
Inertial Navigation ◽  
Calibration Method ◽  
Measurement Information ◽  
Dynamic Calibration ◽  
Navigation Systems ◽  
Inertial Navigation Systems ◽  
Additional Information ◽  
Two Systems ◽  
Dynamic Calibration Method

Generally, in order to ensure the reliability of Navigation system, vehicles are usually equipped with two or more sets of inertial navigation systems (INSs). Fusion of navigation measurement information from different sets of INSs can improve the accuracy of autonomous navigation effectively. However, due to the existence of misalignment angles, the coordinate axes of different systems are usually not in coincidence with each other absolutely, which would lead to serious problems when integrating the attitudes information. Therefore, it is necessary to precisely calibrate and compensate the misalignment angles between different systems. In this paper, a dynamic calibration method of misalignment angles between two systems was proposed. This method uses the speed and attitude information of two sets of INSs during the movement of the vehicle as measurements to dynamically calibrate the misalignment angles of two systems without additional information sources or other external measuring equipment, such as turntable. A mathematical model of misalignment angles between two INSs was established. The simulation experiment and the INSs vehicle experiments were conducted to verify the effectiveness of the method. The results show that the calibration accuracy of misalignment angles between the two sets of systems can reach to 1″ while using the proposed method.

Optimizing the Learning Process of Feedforward Neural Networks Using Lightning Search Algorithm

2016 ◽  
Vol 25 (06) ◽  
pp. 1650033 ◽  
Author(s):  
Hossam Faris ◽  
Ibrahim Aljarah ◽  
Nailah Al-Madi ◽  
Seyedali Mirjalili
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Optimization Problems ◽  
Search Algorithm ◽  
Optimization Technique ◽  
Back Propagation ◽  
Feedforward Neural Networks ◽  
Training Algorithms ◽  
Local Optima ◽  
Local Solutions

Evolutionary Neural Networks are proven to be beneficial in solving challenging datasets mainly due to the high local optima avoidance. Stochastic operators in such techniques reduce the probability of stagnation in local solutions and assist them to supersede conventional training algorithms such as Back Propagation (BP) and Levenberg-Marquardt (LM). According to the No-Free-Lunch (NFL), however, there is no optimization technique for solving all optimization problems. This means that a Neural Network trained by a new algorithm has the potential to solve a new set of problems or outperform the current techniques in solving existing problems. This motivates our attempts to investigate the efficiency of the recently proposed Evolutionary Algorithm called Lightning Search Algorithm (LSA) in training Neural Network for the first time in the literature. The LSA-based trainer is benchmarked on 16 popular medical diagnosis problems and compared to BP, LM, and 6 other evolutionary trainers. The quantitative and qualitative results show that the LSA algorithm is able to show not only better local solutions avoidance but also faster convergence speed compared to the other algorithms employed. In addition, the statistical test conducted proves that the LSA-based trainer is significantly superior in comparison with the current algorithms on the majority of datasets.

Accelerating GPU Message Communication for Autonomous Navigation Systems

2021 ◽  
Author(s):  
Hao Wu ◽  
Jiangming Jin ◽  
Jidong Zhai ◽  
Yifan Gong ◽  
Wei Liu
Keyword(s):  
Autonomous Navigation ◽  
Navigation Systems ◽  
Message Communication

scholarly journals Optical Recognition of Handwritten Logic Formulas Using Neural Networks

Electronics ◽  
2021 ◽  
Vol 10 (22) ◽  
pp. 2761
Author(s):  
Vaios Ampelakiotis ◽  
Isidoros Perikos ◽  
Ioannis Hatzilygeroudis ◽  
George Tsihrintzis
Keyword(s):  
Neural Networks ◽  
Character Recognition ◽  
Gradient Descent ◽  
Feedforward Neural Networks ◽  
Stochastic Gradient ◽  
Stochastic Gradient Descent ◽  
Training Algorithms ◽  
Gradient Descent Algorithm ◽  
Two Stages ◽  
And Training

In this paper, we present a handwritten character recognition (HCR) system that aims to recognize first-order logic handwritten formulas and create editable text files of the recognized formulas. Dense feedforward neural networks (NNs) are utilized, and their performance is examined under various training conditions and methods. More specifically, after three training algorithms (backpropagation, resilient propagation and stochastic gradient descent) had been tested, we created and trained an NN with the stochastic gradient descent algorithm, optimized by the Adam update rule, which was proved to be the best, using a trainset of 16,750 handwritten image samples of 28 × 28 each and a testset of 7947 samples. The final accuracy achieved is 90.13%. The general methodology followed consists of two stages: the image processing and the NN design and training. Finally, an application has been created that implements the methodology and automatically recognizes handwritten logic formulas. An interesting feature of the application is that it allows for creating new, user-oriented training sets and parameter settings, and thus new NN models.

Sign in / Sign up

Export Citation Format

Share Document