Hardware Model Research Articles

Intelligent water quality monitoring and domestic usage management system using the internet of things (IoT)

As time goes on, water supplies are becoming increasingly limited and unpredictable. It is critical to know the water's quality. Using an intuitive mobile app interface, this project integrates the features needed to give a reliable method of measuring water quality and use. The ESP-8266 Node MCU is used as the microcontroller in the hardware model. Data may be transmitted over the internet using this module's network connectivity. The temperature of the water is measured using a temperature sensor. The temperature sensor is DS18B20, which is being utilized due to the availability of dependable library files. The pH sensor was designed with the particular purpose of determining the liquid's pH. You may use the flow sensor to figure out how much water you've used. Using Google Firebase, sensors can send and store data to and from the cloud. Firebase was chosen above other cloud providers because of its reputation for speed and security. It is possible to view the output from any location through the internet using an Android mobile application. Using the app, you can keep tabs on the sensed information from wherever you happen to be at any given time. When a user's use limit is exceeded, the app notifies them. Users will be able to see the actual number of liters of water they've used. To put it another way, the current systems are expensive and bulky. This system is a straightforward collection of components that may be sold as a finished good. Energy usage of the system has been lowered and a secure mobile app has been developed to show data.

Modeling of Packet Error Rate Distribution Based on Received Signal Strength Indications in OMNeT++ for Wake-Up Receivers.

Wireless sensor network (WSN) with energy-saving capabilities have drawn considerable attention in recent years, as they are the key for long-term monitoring and embedded applications. To improve the power efficiency of wireless sensor nodes, a wake-up technology was introduced in the research community. Such a device reduces the system's energy consumption without affecting the latency. Thereby, the introduction of wake-up receiver (WuRx)-based technology has grown in several sectors. The use of WuRx in a real environment without consideration of physical environmental conditions, such as the reflection, refraction, and diffraction caused by different materials, that affect the reliability of the whole network. Indeed, the simulation of different protocols and scenarios under such circumstances is a success key for a reliable WSN. Simulating different scenarios is required to evaluate the proposed architecture before its deployment in a real-world environment. The contribution of this study emerges in the modeling of different link quality metrics, both hardware and software metrics that will be integrated into an objective modular network testbed in C++ (OMNeT++) discrete event simulator afterward are discussed, with the received signal strength indicator (RSSI) for the hardware metric case and the packet error rate (PER) for the software metric study case using WuRx based on a wake-up matcher and SPIRIT1 transceiver. The different behaviors of the two chips are modeled using machine learning (ML) regression to define parameters such as sensitivity and transition interval for the PER for both radio modules. The generated module was able to detect the variation in the PER distribution as a response in the real experiment output by implementing different analytical functions in the simulator.

Analysis of transmission line modeling routines by using offsets measured least squares regression ant lion optimizer in ORPD and ELD problems

This paper proposed an offset measured least regression based ALO to solve ORPD and ELD problems of IEEE 57 bus system designed with different transmission line models. These two problems are highly non-linear and non-convex defiance optimization of problem. The solution of ALO depends on exploration and exploitation if the difference between local and global variables is large, therefore chance to miss the best optimal solution. The weighted elitism phase of the algorithm gives diversified results because exploration is more biased toward elite particles. Which is due to decreasing of random walk to achieve the convergence characteristics. The proposed LSR-EALO can balance both exploration and exploitation, which improves the solution of optimization problem. Simulation is performed with proposed method on different IEEE 57 bus power system models, such as the positive sequence, 3-Phase PI, and distributed CP transmission lines based power systems, and lumped PI lines based low voltage hardware model (LVHM). In this paper, the ORPD problem was used to describe control variables like generator voltage, tap changers of transformers, and switching of capacitor banks subjected to power loss minimization function. Also, described voltage deviation and voltage stability index. Similarly, the ELD was described the active power allocation among generators to meet the sum of load demand and losses in the systems at minimum fuel cost function. And in depth analysis of the optimization results shows accuracy of control variables in ORPD and ELD problems. Also, the effectiveness of proposed method was also verified by comparing results with other meta heuristic algorithms.

ПРОГРАММНО-АППАРАТНОЕ МОДЕЛИРОВАНИЕ КОМПЛЕКСИРОВАННОЙ ИНЕРЦИАЛЬНОЙ НАВИГАЦИОННОЙ СИСТЕМЫ НАЗЕМНОГО ОБЪЕКТА

The purpose of this paper is creation of an efficient software and hardware model of a groundmobile object navigation system. Simulation process is one of the key instruments for engineering solutionsdevelopment at all stages of a complex technical system lifecycle. Navigation system model describedin this paper is software-hardware by nature and is implemented as software modules that supporthardware interaction with each other. The described system of ground object navigation systemsimulation has been developed in the course of several design projects and scientific research works.The developed model of a ground mobile object navigation system includes several programming units.Each of these units simulates the operation process as close as possible to the original. Internal andexternal interaction interfaces and navigation system communication protocols are also simulated withthe account for real object operation cyclograms. It should be noted that one of the main advantages ofa software simulation over a developmental testing or breadboarding is the opportunity to perform aprofound analysis of operation algorithms without the use of any additional resources. At the same time,it is often necessary to perform firmware design in order to make the simulation process identical withthe real operation. The process of navigation system architecture simulation allows performing a comparativeanalysis of several base components sets and their interaction patterns. Such comparison resultsin an optimal system structure providing the information user with navigation data in the mostefficient way. The paper determines basic simulation process tasks. The simulation stage may becomecrucial for complex systems design as it may aid achieving the best technical result. The proposed structure,key elements of a ground navigation simulation system and approaches to their interaction arrangementhave been approved in the course of several research and development works. At the sametime, it was found that the validity of developed models and their compliance to simulated units allowsperforming a highly reliable navigation system analysis. The developed software-hardware model allowedprocessing telemetry records obtained under different product operation conditions, performingtheir analysis and developing engineering solutions that increase the efficiency of navigation systems.

Reverse Power Protection of an Alternator

Power plant generators are important components of an electrical energy system. They should be constantly monitored and protected in order to maintain the quality and reliability of the power supply. Otherwise generator may operate incorrectly resulting in poor power quality and instability during events like faults, equipment failure, reverse power flow(RPF) etc., the presence of large distributed energy resources is the major cause to the reverse power flow during low loads/excess generation at distribution level. If reverse power flow (RPF) is not resolved, it leads to incorrect operation of generators or affect connected devices. Therefore it is necessary to monitor and prevent reverse powerflow condition.This paper dealswith reversepower flow protection, here hardware model representing generator connected to the grid is developed using low rating permanent magnet motors, sensors, PIC16F877A microcontroller, regulated power supply, relay driver, Liquid Crystal display (LCD) and relay, to monitor and protect generator from RPF. Key Words: Reverse power flow, Permanent magnet motors, PIC16F877A Microcontroller, Changeover & protection relay.

Design of smart autonomous solar panel with cascaded SEPIC-boost converter for high voltage renewable applications

ABSTRACT The most admired alternative to conventional energy sources is solar energy. In north-eastern regions of India, where availability of sun is unpredictable but irradiance level is high, special observation is needed to upgrade the solar collection systems. The main objective of the research work is to develop a renewable system, which can extract maximum voltage using dual axis movable surface and design a cascaded converter, which can further increase the voltage while maintaining high efficiency. The output of the smart independent self-regulating solar panel hardware model proposed in this paper is fed into a DC–DC cascaded Boost-SEPIC system for renewable-based applications. The newly created renewable system is built to move the panel so that the most voltage is extracted. The DC–DC cascaded converter receives power from the solar panel’s output. To assess the system’s effectiveness and voltage gain, a thorough circuit analysis is carried out. The converter is tested using various duty cycles. With a 50% duty ratio and 90% efficiency, the suggested converter produces a 40 V DC output from a 20 V DC input voltage from a solar panel. Results from an experimental prototype are used to verify the converter’s viability.

Design and Modeling of Modified Interleaved Phase-Shifted Semi-Bridgeless Boost Converter for EV Battery Charging Applications

Electric vehicles (EVs) are set to become one of the domestic transportation systems that are highly preferred over conventional vehicles. Due to the huge demand for and cost of fuel, many people are switching over to EVs. Companies such as Tesla, BMW, Audi, and Mercedes have started marketing EVs. These EVs need charging stations to charge the batteries. The challenges for EV batteries require the implementation of features such as fast charging, long-run utilization, reduced heat emission, a light weight, and a small size. However, fast charging using conventional converters generates an imbalance in current injection due to the passive component selection. In this study, a converter is proposed that uses an interleaved network that provides a balanced current injection; i.e., an improved interleaved phase-shifted semi-bridgeless boost converter (IIPSSBBC) is designed for EV battery charging applications. The suggested approach is mathematically designed using MATLAB/Simulink (2021) software. The result shows that the battery charging current achieves about 16.5 A, which is relatively more than conventional systems. Moreover, the charging time of the proposed converter is about 6 hrs for a 50 Ah battery with a discharge load capacity of 5000 W, which is relatively less than the conventional method. In a nutshell, compared with conventional converters, the IIPSSBBC performs better, and, notably, the charging speed and current injection are increased by two times the amount. Further, a prototype hardware model is developed to assess the performance of the proposed converter.

Simultaneously Transmitting and Reflecting (STAR)-RISs: Are They Applicable to Dual-Sided Incidence?

A hardware model and a signal model are proposed for dual-sided simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RISs), where the signal simultaneously incident on both sides of the surface. Based on the proposed hardware model, signal models for dual-sided STAR-RISs are developed. For elements with scalar surface impedance, it is proved that their transmission and reflection coefficients on both sides are identical. Based on the obtained symmetrical dual-sided STAR model, a STAR-RIS-aided two-user uplink communication system is investigated for both non-orthogonal multiple access (NOMA) and orthogonal multiple access (OMA) schemes. Analytical results for the outage probabilities for users are derived in the high transmit signal-to-noise ratio (SNR) regime. Numerical results demonstrate the performance gain of NOMA over OMA and reveal that the outage probability error floor can be lowered by adjusting the ratio between the amplitudes of transmission and reflection signals.

Modeling and Simulation of Circuit-Level Nonidealities for an Analog Computing Design Approach With Application to EEG Feature Extraction

This article presents a design approach for the modeling and simulation of ultralow power (ULP) analog computing machine learning (ML) circuits for seizure detection using electroencephalography (EEG) signals in wearable health monitoring applications. In this article, we describe a new analog system modeling and simulation technique to associate power consumption, noise, linearity, and other critical performance parameters of analog circuits with the classification accuracy of a given ML network, which allows to realize a power and performance optimized analog ML hardware implementation based on diverse application-specific needs. We carried out circuit simulations to obtain nonidealities, which are then mathematically modeled for an accurate mapping. We have modeled noise, nonlinearity, resolution, and process variations such that the model can accurately obtain the classification accuracy of the analog computing-based seizure detection system. Noise has been modeled as an input-referred white noise that can be directly added at the input. Device process and temperature variations were modeled as random fluctuations in circuit parameters, such as gain and cut-off frequency. Nonlinearity was mathematically modeled as a power series. The combined system level model was then simulated for classification accuracy assessments. The design approach helps to optimize power and area during the development of tailored analog circuits for ML networks with the ability to potentially trade power and performance goals while still ensuring the required classification accuracy. The simulation technique also enables to determine target specifications for each circuit block in the analog computing hardware. This is achieved by developing the ML hardware model, and investigating the effect of circuit nonidealities on classification accuracy. Simulation of an analog computing EEG seizure detection block shows a classification accuracy of 91%. The proposed modeling approach will significantly reduce design time and complexity of large analog computing systems. Two feature extraction approaches are also compared for an analog computing architecture.

PROBLEMS OF THE SYNTHESIS OF NETWORK EQUIPMENT MODELS AND METHODS OF THEIR SOLUTION

The research examines the problems of the development of software-configured networks (PSNs) and focuses on hybrid software-configured equipment (PKO). The relevance of the research is emphasized in the context of modern requirements for network technologies and certain difficulties in modeling such devices. A comparative analysis of simulation modeling tools of computer networks was carried out, the advantages and disadvantages of various solutions were determined. We chose to use OMNeT++ to implement a simulation-analytical model of hybrid software-configurable hardware due to its open source, modular architecture, and available documentation. The work describes and defines the problems of synthesis of network equipment models, taking into account the differences in the architecture and configuration of hybrid software-configured devices. It is noted that many existing modeling systems do not have ready-made models that adequately reflect the probabilistic-temporal characteristics of the internal processes of these devices. The purpose of the research is given, which consists in the development of methods and algorithms for the automated synthesis of a simulation-analytical model of hybrid software-configured equipment for the study of packet delay in telecommunication networks. Five tasks are set, including the development of a model, methods and algorithms for automated experimental research and model synthesis. The choice of OMNeT++ is determined by its characteristics, and the model synthesis process itself is defined as a complex task that takes into account the configurational diversity of the equipment. A review of the main publications highlighting the problems of the development of software-configurable devices is carried out, and the expediency of using simulation modeling for the study of hybrid software-configurable devices, as well as challenges and prospects for further research, is indicated.

An Exploration of State-of-the-Art Automation Frameworks for FPGA-Based DNN Acceleration

FPGA-based acceleration is considered a promising approach to improve the performance and power efficiency of Deep Neural Network (DNN) inference tasks. However, mapping a DNN onto an FPGA is not trivial. To make this easier, various automation frameworks have been proposed. Among them, FINN and Vitis AI, both developed by Xilinx, are two key players. They represent two different philosophies in designing FPGA-based DNN accelerators: dataflow-style and overlay-style architectures. Dataflow architectures are generally expected to provide better performance and power efficiency but have a major drawback in that they scale very poorly to the size of the target DNN. Advanced frameworks like FINN alleviate this drawback by transforming the target DNN into operations that can be time-multiplexed on fewer hardware resources. This approach, however, is challenging because of the difficulty in transforming the target DNN and raises a question as to whether the generated dataflow architectures retain a significant advantage over overlay architectures in terms of performance and power efficiency. This paper aims to clarify it by conducting an in-depth exploration of FINN and Vitis AI. For this purpose, we extend the FINN’s development flow to be able to use the same target hardware and DNN model to evaluate each framework. We demonstrate the effectiveness of the FPGA-based acceleration by providing a comparison with two reference platforms: an NVIDIA Jetson Nano Developer Kit with a similar power budget to our target FPGA hardware, and a high-performance desktop computer with an Intel Core i7-11700K CPU. The results show that despite the use of the time-multiplexing approach, the FINN-based accelerator can still outperform the Vitis-AI-based accelerator by a significant margin, 8.4x in terms of latency, 3.0x in terms of throughput, and 3.3x in terms of power efficiency. The outcome of the comparison with the NVIDIA Jetson Nano Developer Kit and the desktop computer is also overwhelmingly favorable to the FINN-based accelerator. This indicates that, even in the case of using automation frameworks, DNN accelerators on FPGAs can still yield significant performance and power efficiency gains compared with GPUs and CPUs.

A HYBRID QUANTUM-PERFECTED MODEL OF ARTIFICIAL INTELLIGENCE IN THE PROBLEM OF AUTOMATIC RECOGNITION AND FAST CONVERSION OF UNSTRUCTURED TEXT INFORMATION INTO SPATIAL

Background. Efficiently converting large amounts of unstructured text data into spatial information is crucial for managing water distribution systems. This allows for the conversion of extensive sets of text information, such as reports, orders, letters, and other documents, into point classes of spatial objects in geographic information systems. To tackle this challenge, a promising new approach involves combining hybrid quantum-classical neural networks with geo-information technologies. Methods. The study utilized quantum-enhanced hybrid neural networks in combination with GIS methods to identify named entities such as personal accounts and balance sheet objects of Kyivvodokanal by their addresses and geocoding. This information was then published on a geoportal using the ArcGIS Enterprise platform in real-time, which holds great promise for effective water management. The performance of the developed model was evaluated by accuracy indicators, recall parameters, and weighted harmonic average of accuracy and recall. Results. The obtained results indicate that the developed hybrid quantum-classical model of artificial intelligence can be successfully applied to transform large volumes of unstructured textual information into spatial information. The model was integrated into GIS using ArcGIS Enterprise. By combining the obtained point classes of spatial objects with already existing data, methods of spatial connections, an interactive map with an update interval of every five minutes was developed. Conclusions. Taking advantage of quantum computing and combining it with classical hardware and classical AI models, it became possible to achieve similar and even better performance in various tasks compared to state-of-the-art methods. Quantum natural language processing is a promising new field that has the potential to revolutionize the way one analyzes and understands human language.

Aging or Glitching? What Leads to Poor Android Responsiveness and What Can We Do About It?

Almost all Android users have ever experienced poor responsiveness, including the common frame dropping events—slow rendering (SR) and frozen frames (FF), as well as the uncommon Application Not Responding (ANR) and System Not Responding (SNR) that directly disrupt user experience. This work takes two complementary approaches, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">controlled benchmarking</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">in-the-wild crowdsourcing</i> , to comprehensively understand their prevalence, characteristics, and root causes, which turn out to be significantly different from common understandings and prior studies. We find that SR, FF, ANR, and SNR all occur prevalently on all the studied hardware models of Android phones, and better hardware does not seem to relieve ANR/SNR. Most surprisingly, they are oftentimes ascribed to defective software design that incurs substantial resource overuse—lightweight apps can experience severe SR/FF events due to <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">redundant UI rendering</i> , and the most ANR/SNR events stem from Android's aggressive implementation of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">write amplification mitigation</i> . In fact, the former can be effectively overcome by simplifying the apps' UI hierarchy, and we design a practical approach to address almost all ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$&gt;$</tex-math></inline-formula> 99%) of the latter while only decreasing 3% of the data write speed with large-scale deployment. We have released our measurement code/data to the research community.

An overview of smart grid smart meters with fault detection

Smart meters (SM) arrangement play an essential role in Smart Grid (SG), serving as advanced energy meters crucial for monitoring and managing energy consumption. This paper provides an in-depth exploration of the significance, functions, and technological aspects of smart meters within the context of the SG. Smart meters acquire data from endusers' load devices, offering a comprehensive measurement of consumers' energy consumption. The obtained information is then communicated to utility companies and system operators, facilitating efficient energy management. The integration of control devices and a number of sensors, supported by a dedicated communication infrastructure, enhances the reliability and functionality of smart meters. The paper summarizes hardware model key facets and functions of smart meters, offering a comprehensive view of their capabilities.

Low Communication Secure Computation From Semi-Trusted Hardware

In privacy-preserving machine learning and many other applications, the involved parties want to obtain the computation result without revealing their private inputs. Secure computation aims to solve this problem, but current secure computation protocols often fail to provide efficient solutions due to large communication, especially in a real-life Internet network where the bandwidth and the delay can be unsatisfying. Assuming the existence of a trusted hardware component that is resilient to side-channel attacks and will faithfully compute a pre-agreed program, secure computation can be realized by each party sending its input to the hardware and receiving the execution result back. However, a recent work of Lu <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">et al</i> . (ESORICS’21) points out that the hardware components can’t be fully trusted. In this work, we improve the semi-trusted hardware model of Lu <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">et al</i> ., and we propose secure computation protocols with low communication in the new model. We observe that the ESORICS’21 two-party computation protocol have some security flaws; in this work, we fix them and improve its online efficiency. Moreover, we propose an efficient constant-round secure multi-party computation protocol which has a communication cost of ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</i> – 1)λ + 2( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</i> – 1)ℓ bits, where <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</i> is the number of the parties, λ is the security parameter and ℓ is the input/output size. The computation cost of our multi-party protocol is also much smaller than current best-known constant-round protocols.

ARCADE PI THEME: RETRO DIGITAL GAMES

Currently for the most nostalgic in the world of games, playing a game title on the Super Nintendo on an original console is practically a rarity. There are alternatives on the internet to play on computers, but it is always necessary to take a computer or notebook to enjoy these games and we are not always able to take them to other places. This project intends to demonstrate how a classic game can be digitally preserved and show this retro gaming culture to younger people. All through a mini- microcomputer known as Orange PI - old digital games. It will address the functionality strategies and configurations of the system that will be installed on the console, hardware and software model and its components and equipment for such development. Soon after the introduction, some authors thoughts on this project will be shown, their contribution to having a more in-depth look at classic games, which helped a lot in the development of the project.Its methodology details the construction of the Orange PI - old digital games, all these steps will be described in detail from its assembly to the allocation of the games library in the console. Its result was very satisfactory, for what was wanted to develop, but this console can still be improved a lot.

Hardware Trojan Detection and Mitigation in NoC using Key authentication and Obfuscation Techniques

Today's Multiprocessor System-on-Chip (MPSoC) contains many cores and integrated circuits. Due to the current requirements of communication, we make use of Network-on-Chip (NoC) to obtain high throughput and low latency. NoC is a communication architecture used in the processor cores to transfer data from source to destination through several nodes. Since NoC deals with on-chip interconnection for data transmission, it will be a good prey for data leakage and other security attacks. One such way of attacking is done by a third-party vendor introducing Hardware Trojans (HTs) into routers of NoC architecture. This can cause packets to traverse in wrong paths, leak/extract information and cause Denial-of-Service (DoS) degrading the system performance. In this paper, a novel HT detection and mitigation approach using obfuscation and key-based authentication technique is proposed. The proposed technique prevents any illegal transitions between routers thereby protecting data from malicious activities, such as packet misrouting and information leakage. The proposed technique is evaluated on a 4x4 NoC architecture under synthetic traffic pattern and benchmarks, the hardware model is synthesized in Cadence Tool with 90nm technology. The introduced Hardware Trojan affects 8% of packets passing through infected router. Experimental results demonstrate that the proposed technique prevents those 10-15% of packets infected from the HT effect. Our proposed work has negligible power and area overhead of 8.6% and 2% respectively.

Quantum-inspired algorithm for direct multi-class classification

Over the last few decades, quantum machine learning has emerged as a groundbreaking discipline. Harnessing the peculiarities of quantum computation for machine learning tasks offers promising advantages. Quantum-inspired machine learning has revealed how relevant benefits for machine learning problems can be obtained using the quantum information theory even without employing quantum computers. In the recent past, experiments have demonstrated how to design an algorithm for binary classification inspired by the method of quantum state discrimination, which exhibits high performance with respect to several standard classifiers. However, a generalization of this quantum-inspired binary classifier to a multi-class scenario remains nontrivial. Typically, a simple solution in machine learning decomposes multi-class classification into a combinatorial number of binary classifications, with a concomitant increase in computational resources. In this study, we introduce a quantum-inspired classifier that avoids this problem. Inspired by quantum state discrimination, our classifier performs multi-class classification directly without using binary classifiers. We first compared the performance of the quantum-inspired multi-class classifier with eleven standard classifiers. The comparison revealed an excellent performance of the quantum-inspired classifier. Comparing these results with those obtained using the decomposition in binary classifiers shows that our method improves the accuracy and reduces the time complexity. Therefore, the quantum-inspired machine learning algorithm proposed in this work is an effective and efficient framework for multi-class classification. Finally, although these advantages can be attained without employing any quantum component in the hardware, we discuss how it is possible to implement the model in quantum hardware.

Machine vision system for real-time debris detection on mechanical wild blueberry harvesters

Improving fruit quality is important to wild blueberry producers. Plant debris like leaves and stems caused a reduction in fruit quality during mechanical harvesting. A debris detection system was developed using two Logitech C920 webcam cameras mounted before and after the blower fan on a commercial mechanical wild blueberry harvester. Images were collected from two commercially managed wild blueberry fields located in central Nova Scotia to develop a dataset of 1000 images. Two convolutional neural networks (CNNs), YOLOv3 and YOLOv3-SPP, were used and compared for performance analysis. The CNNs were trained using no augmented images, images with five augmentations (sharpening, brightness, contrast, gamma correction, and saturation), and images with five augmentations that were more heavily weighted towards gamma. Four different computer hardware packages (Hewlett-Packard, Shuttle XPC, Jetson TX2, and GPU-based Desktop) were used to determine the appropriate hardware and CNN model combination for real-time performance. YOLOv3 and YOLOv3-SPP were trained to mAP scores of 72.87 and 74.38% using the dataset with more gamma-augmented images. This was an improvement over the dataset with no augmentation (67.75; 68.49%) and equally numbered augmentations (71.26%; 73.03%). YOLOv3-SPP combined with an Intel® Core™ i9-7900X CPU and Nvidia GeForce RTX™ 2080 Ti GPU in a desktop computer achieved the fastest detection rate of 33.30 ms. The detection speed on all other hardware configurations exceeded 33.30 ms, indicating that a powerful desktop GPU is required for real-time performance of this task. With adequate processing hardware, the developed technology could be integrated into a control system to automatically adjust brushes based on conveyor information from mechanical wild blueberry harvesters to make cleaning berries easier and enhance fruit quality.

Design and hardware demonstration of smart meter by cloud interface

AbstractIn this work, the proposed model is developed by employing a smart metre design which is done by controlling and monitoring the system frequency. This model estimates the changes in the frequency in accordance to the loading conditions of the power system, overload or under load‐conditions respectively. The estimated frequency is analysed by employing a smart metre design, which estimates the change in the system frequency caused by overload or under‐load conditions and compared with a reference frequency value set by the load dispatching unit in the control server. In this analysis, the line responsible for the frequency change are being isolated from the rest of the system or given only based on the demand power as per the priority loads. The proposed model is equipped with an embedded realistic system set along with a synchronised network and central server by using a cloud computing approach as a test bed laboratory set up. The parameters of the electric power are based on load forecasting involved for setting the required reference frequency. The model is developed with a realistic approach by developing the prototype. The work employs both software and hardware modelling with cloud interface. A complete hardware demonstration rig is developed with a smart metre design and experimental results are studied and demonstrated using cloud interface.