Showing papers presented at "International Symposium on Industrial Electronics in 2021"

VT-ADL: A Vision Transformer Network for Image Anomaly Detection and Localization

Abstract: We present a transformer-based image anomaly detection and localization network. Our proposed model is a combination of a reconstruction-based approach and patch embedding. The use of transformer networks helps preserving the spatial information of the embedded patches, which is later processed by a Gaussian mixture density network to localize the anomalous areas. In addition, we also publish BTAD, a real-world industrial anomaly dataset. Our results are compared with other state-of-the-art algorithms using publicly available datasets like MNIST and MVTec.

A GIS-based Optimal Facility Location Framework for Fast Electric Vehicle Charging Stations

Abstract: Deeper decarbonization of the transport sector requires building a wide coverage electric vehicle charging network that can meet driver's mobility patterns and refueling habits in a seamless manner. Currently, major market players mainly deploy chargers at existing public parking spaces at hotels, shopping centers, etc. On the other hand, gas/petroleum retail business is a century-old industry and “optimized” to serve the refueling needs of the drivers and they come to the forefront as “good” locations to site chargers. To that end, this paper addresses the fast charging station location problem in an urban environment. The optimization problem is formulated as a maximum coverage location problem (MCLP) and existing locations of petrol/fuel stations are considered as candidate locations. Using QGIS software, a geographic information system (GIS) based platform is developed and integrated with a linear-programming relaxation based MCLP algorithm developed in Python. The city of Raleigh, North Carolina with actual geo-spatial data is chosen as a case study. Both census population and highway traffic data are considered as demand metrics to mimic drivers without dedicated chargers and vehicles on highways who need a recharge. A number of evaluations are performed to explore the trade-off between the number of locations and the physical coverage space. Furthermore, comparative analysis show that locating fast chargers in existing petrol stations improve demand coverage by more than 50 % when compared to existing fast charging station locations.

Isolated MMC-based ac/ac stage for ultrafast chargers

Abstract: The application of indirect and direct modular multilevel converter-based topologies in an isolated ultrafast charger, operating from a three-phase medium-voltage grid, is compared. The most promising circuit is an isolated direct ac/ac modular multilevel converter, in which four-quadrant operation is possible by employing full-bridge sub-modules. The direct ac/ac conversion reduces the cost and volume of fast charging stations by eliminating line frequency transformers. An analytical model focusing on the decomposition of circuit states is employed to present the control scheme. Subsequently, the approach is validated with simulation results, corroborating the suitability for high-power bidirectional battery chargers.

Multiple Object Trackers in OpenCV: A Benchmark

Abstract: Object tracking is one of the most important and fundamental disciplines of Computer Vision. Many Computer Vision applications require specific object tracking capabilities, including autonomous and smart vehicles, video surveillance, medical treatments, and many others. The OpenCV as one of the most popular libraries for Computer Vision includes several hundred Computer Vision algorithms. Object tracking tasks in the library can be roughly clustered in single and multiple object trackers. The library is widely used for real-time applications, but there are a lot of unanswered questions such as when to use a specific tracker, how to evaluate its performance, and for what kind of objects will the tracker yield the best results? In this paper, we evaluate 7 trackers implemented in OpenCV against the MOT20 dataset. The results are shown based on Multiple Object Tracking Accuracy (MOTA) and Multiple Object Tracking Precision (MOTP) metrics.

Human-Cyber-Physical System Integration (HSI) in Industry 4.0: design and evaluation methods

Abstract: This overview article describes the design and use space of integrating humans and cyber-physical systems in Industry 4.0, with special regards to the interplay of analysis, design and evaluation methods and phases. Starting with an introduction into the challenges of Industry 4.0 and an overview on existing methods of system design, the design and use space of method models is described and exemplified with examples from Industry 4.0. An extended U-Method of iterative analysis, design and evaluation is derived, described in theory and exemplified with practical examples. An outlook identifies potential roadmaps of future design and evaluation methods especially for industry 4.0.

Fault Diagnosis of Rotating Machinery based on Domain Adversarial Training of Neural Networks

Abstract: With the increased requirement of reliable facility operations of rotating machinery, the prediction and diagnosis of fault signals are crucial to improve the safety of equipment. Fault diagnosis with artificial intelligence is an effective method to classify the machinery failure rapidly and automatically. However, the training process requires mass of labeled data which is impractical to obtain. Transfer learning are promoted to overcome the shortage of data by transferring the results of related study and combining current resources to diagnosis. Domain adversarial training of neural networks (DANN) as a typical model of transfer learning efficiently solves this problem. In addition, cohesion evaluation technique is used in the data preprocessing to establish low-dimensional sensitivity feature vectors. In order to verify the effectiveness of the methods, experiments are conducted on two different platforms for transfer learning. The experiment reveals that the proposed method can achieve better results than conventional methods under several evaluation metrics.

Feature-based Egocentric Grasp Pose Classification for Expanding Human-Object Interactions

Abstract: This paper presents a framework for classifying human hand pose, especially in grasping object intuitively. First, we propose a system based on the stereo infra-red image as a sensor that can produce hand coordinates in 3-dimensional space. We use egocentric vision because it can get uniform and natural data with only a single sensor module. Second, we transformed the position to get the angle information for each joint on the finger. Third, we designed an intelligent system based on Multi-Layer Perceptron (MLP) to process angular data to obtain classification results according to the Cutkosky grasp taxonomy. Finally, we compared the results on several similar objects and evaluated their classification accuracy. In the validation phase, the results yielded an accuracy of 16 grasp pose classification is 89,60%. In real-time testing, the results yielded an accuracy of 81.93%. This result shows feature-based learning can reduce the complexity and training time of the MLP. Furthermore, a small amount of training data is sufficient for the training and implementation.

Efficient GAN - Based Unsupervised Anomaly Sound Detection for Refrigeration Units

Abstract: A smart factory or Industry 4.0 is creating an epoch for manufacturing and its production lines. It reduces the total cost by monitoring and predicting the expected faults of factory lines and products. One of the essential challenges is to develop a technology to detect and predict abnormalities at an early stage without human resources. For this reason, the automation of anomaly detection is now attracting attention. Many statistical and machine-learning methods have been studied for anomaly detection. In this study, we focus on a refrigeration system for large storage, where the failures of the system will cause enormous losses. Moreover, this type of system was independently designed according to the environment, location, and storage items. Under this condition, it is difficult to train discriminative models for anomaly detection using training data that include failure data. In addition, it is indispensable to provide a basis for determining whether the system is abnormal to achieve future treatments. Therefore, deep generative models are used to achieve unsupervised abnormality detection. Because the sensing system's cost for detecting system failures should be reduced, the proposed system uses low-cost microphone arrays to monitor sounds and source locations. The system also provides a rationale by visualizing and mentioning irregular sounds. Furthermore, this study compared various deep generative models in terms of accuracy and showed that the Efficient GAN-based method achieved the highest accuracy.

Using Visible/Near-Infrared Reflectance Spectroscopy and Chemometrics for the Rapid Evaluation of Two Panamanian Watermelon (Citrullus lanatus) Varieties

Abstract: In this study, visible/near-infrared (Vis/NIR) reflectance spectroscopy measurement in the 300-900 nm wavelength range, are used as external parameter to evaluate the internal parameters (Brix content and pH) of two local watermelon varieties, Quetzaly (N=8) and Mickylee (N=6), in a noninvasive manner. To assess this relationship three Chemometrics and machine learning methods were used: Principal Component Regression (PCR), Partial Least Squares (PLS) and Multilayer Perceptron Regression (MLPR). Spectral Data augmentation was used to generate more observations from the original spectra. Our results suggest Quetzaly has a higher Brix content and that all three methods can be used to predict it with confidence. However, PLS shows a higher correlation coefficient of 0.98 and standard errors of prediction of ≤0.1, closely followed by PCR with 0.96 and errors of ≤0.01 and ≤0.04, respectively. In the case of pH all methods produce good results, with a PLS achieving correlation coefficient of 0.91-0.94 with errors of prediction on the <0.05; and lower results for the other two methods. This study demonstrates the value of non-invasive Vis/NIR Reflectance spectroscopy as quality assessment in a watermelon classification system. More importantly, this system can be later implemented in local watermelon fruit packing centers in Panama. This can also be used in other export fruits or vegetables, helping local producers.

Data-driven Thermal Anomaly Detection for Batteries using Unsupervised Shape Clustering

Abstract: For electric vehicles (EV) and energy storage (ES) batteries, thermal runaway is a critical issue as it can lead to uncontrollable fires or even explosions. Thermal anomaly detection can identify problematic battery packs that may eventually undergo thermal runaway. However, there are common challenges like data unavailability, environment and configuration variations, and battery aging. We propose a data-driven method to detect battery thermal anomaly based on comparing shape-similarity between thermal measurements. Based on their shapes, the measurements are continuously being grouped into different clusters. Anomaly is detected by monitoring deviations within the clusters. Unlike model-based or other data-driven methods, the proposed method is robust to data loss and requires minimal reference data for different pack configurations. As the initial experimental results show, the method not only can be more accurate than the onboard BMS and but also can detect unforeseen anomalies at the early stage.

Recurrent Online and Transfer Learning of a CNC-Machining Center with Support Vector Machines

Abstract: Data-driven learning methods represent a promising field of research to complement classical approaches in the area of control theory. Within the German cluster of excellence “Internet of Production” (IoP), model-based control strategies are researched using collective knowledge accumulated in shared databases, and adapted online according to sensor acquired data. With their inherent generalization ability and affinity for greybox modeling, Support Vector Machines (SVM) are very suitable for such online identification and adaption. However, the computational efficiency of the identification, while maintaining accuracy, is crucial for the real-time capability of the overall framework. This work compares different definitions of the learning problem with SVM for the identification of dynamic systems. Computational efficiency within the given framework is thereby of particular interest. In addition, an extension of existing libraries by transfer learning capabilities is investigated to further speed up the recurrent online identification scheme. The results suggest that SVM with “Sequential Minimal Optimization” (SMO) qualify as a real-time capable general purpose identification approach for model-based control of the derived framework. The addition of transfer learning heavily contributes to the real-time capability.

Obtaining Object Information from Stereo Vision System for Autonomous Vehicles

Abstract: Autonomous navigation is a task that requires the most information it can get from its surroundings for safety, accuracy and decision making. This paper presents a stereo vision system capable of obtaining object information for autonomous vehicles. The stereo vision system is merged with an object detection algorithm to carry out this task. The merge between these systems is a proposal to create a vision system for autonomous navigation tasks that provides information about objects of interest in the field of view, such as object description, distance, height and width.

Modeling and Control of Two-mass Resonant System Based on Element Description Method

Abstract: This paper introduces a semi-automatic modeling and controller design method of vibration system based on an element description method. Generally, the controller of a vibration system is designed by approximating the control target to a multi-mass resonant system. In this study, the vibration system is semi-automatically approximated to a two-mass resonant system by the element description method. After that, the controller is semi-automatically derived based on the approximation. The results of modeling and control generation by the element description method have clear physical meanings and can be explicit knowledge. Experiments with beams and weights confirm the validity of this design framework. From the experimental results, it is confirmed that the controller of the resonant ratio control is automatically generated, and the controller capable of interpreting the physical meaning can be automatically generated by the element description method.

Droop Coefficient Design in Droop Control of Power Converters for Improved Load Sharing: An Artificial Neural Network Approach

Abstract: In this paper, a new approach for the design of the droop coefficient in the droop control of power converters using the artificial neural network (ANN) is proposed. In the first instance, a detailed more electric aircraft (MEA) electrical power system (EPS) circuit model is simulated in a loop using different combinations of the converters droop coefficients within a design space. The inaccurate output DC currents sharing of the converters due to the influence of the unequal cable resistance are then obtained from each of the simulations. The data generated is then used to train the NN to be a dedicated surrogate model of the detailed MEA EPS simulation. Thus, for any user-defined desired current sharing among the converters that are within the design space, the proposed NN can provide the optimal droop coefficients. This NN approach has been verified through simulations to ensure accurate current sharing between the converters as desired. Hence, can be used in the design of the droop coefficient to enhance the performance of the conventional droop control method.

Charging Area Extension of Wireless Fast Charger Based On Relay Coil

Abstract: Wireless faster chargers have been well developed by the various manufacturers (like Xiaomi, Huawei, and Oppo) in the past years. Considering the comparability, the standard Qi-based coils are usually used and will naturally suffer from the misalignment issue. The high-efficiency power transfer can only happen under good alignment. Even a small coil misalignment causes variation in self and mutual inductance, resulting in the drop of output power and transfer efficiency. Instead of customizing a specific transmitter, this paper provides a simple charging area extension method to fully utilize the capability of existing Qi coil, such that the system compatibility can be maintained. The proposed transmitter can still support the charging of Qi-standard based devices, and a replay coil is added outside the small Qi-coil to extend the charging area. Through the circuit-model analysis, this paper explore the power transfer and efficiency characteristics when the replay coil is activated. In the experiment, it shows the proposed charger delivers 50-W power under the standard voltage and current limitation, while the lateral offset between the transmitting coil and receiving coil is 30 mm.

Ricardian Contracts for Industry 4.0 via the Arrowhead Contract Proxy

Abstract: Industry 4.0 will require unprecedented degrees of integration across organizational boundaries, which will put new demands on infrastructure for managing agreements between industrial stakeholders. In this paper, we present a system-of-systems architecture for cross-organizational negotiation of Ricardian contracts. We also describe our implementation of it, based on Eclipse Arrowhead, and how it can produce non-repudiable contracts between local clouds, potentially owned by distinct parties. We discuss how our architecture could impact current business paradigms, as well as arguing that our design, in contrast to most solutions based on smart contracts, avoids to deviate significantly from contemporary legal praxis, which should create better opportunity for industry adoption.

Low-Cost Indoor Localization System Combining Multilateration and Kalman Filter

Abstract: Indoor localization systems play an important role to track objects during their life-cycle in indoor environments, e.g., related to retail, logistics and mobile robotics. These positioning systems use several techniques and technologies to estimate the position of each object, and face several requirements such as position accuracy, security, range of coverage, energy consumption and cost. This paper describes a practical implementation of a BLE (Bluetooth Low Energy) based localization system that combines multilateration and Kalman filter techniques to achieve a low cost solution, maintaining a good position accuracy. The proposed approach was experimentally tested in an indoor environment, with the achieved results showing a clear low cost system presenting an increase of the estimated position accuracy by 10% for an average error of 2.33 meters.

Detecting Out-Of-Control Sensor Signals in Sheet Metal Forming using In-Network Computing

Abstract: The ongoing digitization of production enables the collection of increasing volumes of data. These, in turn, allow for data-driven analysis that has the potential for deepening the process understanding by discovering previously unknown connections between process components and parameters. With these opportunities, however, come substantial challenges as current industrial settings are inadequately equipped for handling these large amounts of data. While setting up a local processing infrastructure is challenging, the limited bandwidth within many shop floors as well as their network access also make an upload of all data to external compute capacities infeasible. What is needed are local, process-aware filters that allow for significant data reduction while retaining data of value that can be used for the subsequent analysis. In this paper, we thus propose to leverage In-Network Computing to dynamically detect different states of the physical processes and then filter the sensor values on the data path. Our presented architecture maps the state detection to the switch-local controlplane while fast filtering decisions are performed at line-rate in the dataplane, thus enabling flexible and quick adjustments of the chosen sensor filtering. At the example of a fine-blanking line, we consequently demonstrate that In-Network Computing can sensibly support previously infeasible data analysis techniques in the industrial production landscape.

Resilient Distributed Frequency Estimation for PEVs Coordinating in Load Frequency Regulation Under Cyber Attacks

Abstract: In view of the fast charging and discharging characteristics of the plug-in electric vehicles (PEVs), this paper coordinates PEVs in conventional load frequency control (LFC). To avoid the installation of measurement devices for each PEV, a resilient distributed frequency estimation strategy is presented for PEVs with the frequency measurements taken only at distribution substations. Since the communication links among PEVs are vulnerable to cyber attacks, this strategy integrates a credibility-based intrusion detection unit to distinguish compromised PEVs. Then, a maximal power acquisition algorithm is designed for well-being PEVs, to obtain sufficient power capacity for the frequency regulation. In simulations, we validate the effectiveness of the credibility-based distributed frequency estimation scheme for PEVs under strongly connected interaction topology. Also, we verify the rapidity of LFC with the coordination of well-being PEVs compared with its conventional counterpart.

Improving voltage harmonics forecasting at a wind farm using deep learning techniques

Abstract: The South African renewable energy mix includes wind, solar, hydro and ocean. This energy mix contributes to the nation energy requirements while reducing dependency on fossil fuel and in the process mitigating the emission of green-house gases. Wind power generation is always associated with the generation of voltage harmonics. Precise predictions of the presence of voltage harmonics is of vital importance in order to ensure clean voltage is coupled to the national grid. A total of 8103 voltage harmonics, measured at Jeffreys Bay Wind Farm in the Eastern Cape Province have been used in our experiments. The proposed model would take two steps to extract important features present in the voltage harmonics signals. The mean voltage amplitude is extracted using moving window segmentation. Long short-term memory (LSTM), a deep learning method, is used in the prediction of voltage harmonics generation based on the voltage features extracted. LSTM is a special kind of recurrent neural network (RNN) capable of learning long-term dependencies. For simplicity the model uses one LSTM layer with 128 hidden neurons. 8103 calculated mean values were used as the expected data to train the model in Matlab. The LSTM model could predict the next 3800 sample mean values with low root mean square error (RMSE).

Finite Control Set Model Predictive Control Without Weighting Factors for Common Grounded Five-Level PV Inverter

Abstract: Elimination of leakage currents is a critical issue for transformerless photovoltaic (PV) inverters. Multilevel common grounded (CG) PV inverters have become promising solutions for mitigating leakage currents. However, developing multilevel outputs and injecting high quality currents impose several challenges for classical controllers. Therefore, this paper presents a finite control set model predictive control, FCS-MPC, without weighting factors for the five level CG switched capacitor cells inverter. Moreover, tuning the weighting factors in the classical FCS-MPC requires performing several cases and comparing the different cases to select the best weighting factors. In the proposed FCS-MPC, the problem of weighting factor adjustment of the classical FCS-MPC methods is eliminated through dividing the cost function objectives. Hence, simplified and robust design is achieved using the proposed FCS-MPC method while avoiding long procedures for tuning the weighting factors. Simulation results of the proposed FCS-MPC method are provided to verify its performance.

Distributed Estimation for Discrete Sequential Systems under Binary Sensors

Abstract: This paper is concerned with the distributed state estimation problem for a class of discrete sequential systems (DSSs) with bounded noises, where each subsystem is observed by several binary sensors. An uncertain approach is employed to deal with estimator design only based on little valid information from binary sensors, and a distributed recursive estimation form is given for DSSs. Then, by using the matrix analysis method and the idea of bounded recursive optimization, each optimal gain of the estimator is designed by constructing different self-relative convex optimization problems that can be solved by the standard software packages. Furthermore, the stability is proved under certain conditions. Finally, an illustrative example is given to show the effectiveness of the proposed methods.

Single-Phase partial power unfolding inverter for photovoltaic string application

Abstract: Partial power converters, also know as fractional or differential power converters among other terminologies, is a very attractive concept that has gained an increasing attention during the last decades. Through a rearrangement of the traditional power conversion structures, the converter only processes a fraction of the total power while the rest is directly transmitted to the load. Consequently, a significant reduction in size and cost of the converter is achieved through the use of low-power ratings devices, allowing also higher efficiencies. Partial power converters have been proposed for various applications, such as photovoltaic (PV) systems. In PV applications single-phase grid-connected PV inverter are composed usually by two-stage power conversion, a full power DC-DC converter and a hard-switched high frequency PWM inverter. Although, different DC-DC partial power converter have been proposed as a first stage, most of the existing solutions deals with constant DC-DC regulation voltage, then the DC-AC grid tied inverter stage still requires to perform all the inversion through hard-switched PWM strategy. In this paper a partial power unfolding dc-dc stage is proposed, which generates a rectified sinusoidal current, leaving only the unfolding duty to the inverter stage. As a result, only one hard-switched PWM partial power stage composes the system, increasing the efficiency. The proposed partial power inverter is applied on a 3.3kW PV string inverter. Simulation results are provide to verify the interest of the proposed partial power DC-AC inverter.

Apple Trees Diseases Detection Through Computer Vision in Embedded Systems

Abstract: In this paper, we address the problem of detecting diseases of apple trees. We report on a computer vision method for apple trees leaves segmentation. For this reason we collect the leaves images in the field using a thermal image camera. Data analysis is carried out using Neural Networks (NN) optimized for running on the embedded systems. We perform a comparative study on the embedded systems, embedded systems enriched with the GPU capability, and the PC. We achieved IoU=0.814. Our results demonstrate that the NNs running on the embedded systems is a promising solution for detecting the trees diseases using embedded systems and open up wide vista for its application in precision agriculture.

Optimal Settings of BTB-VSC in Interconnected Power System Using TFWO

Abstract: This paper presents optimal settings of the Back To Back Voltage Source Converters (BTB-VSC) for optimal operation of a power system. The optimization is applied to achieve three objectives which are minimum total power losses, minimum line loading deviation and minimum voltage deviation. The optimization process is applied using Turbulent Flow of Water based Optimization (TFWO) algorithm. The study is applied to the IEEE 68-bus system which is connecting the New England system and New York system. Three scenarios are considered to connect the New York system with New England system using BTB-VSC in comparison with AC lines. The simulation and optimization process are performed using DIgSIENT power factory software application.

Analysis of 3D CAD MESH Simplification Approaches within the Framework of AR Applications for Industrial Assembly Lines

Abstract: Over the last few years, Augmented Reality (AR) has being investigated as a workforce training tool in industrial assembly lines. However, when 3D CAD meshes are used in AR, embedding them in computationally light AR devices is a challenging task, since such models may be highly detailed, involving up to millions of mesh vertices. In such cases, a possible solution would be to replace the original models by some simplified versions. The crucial challenge is to preserve the quality of the models, while ensuring a fluid rendering on the AR devices, at the usual rates of 30-60 frames per second. In this paper, we analyze different simplification techniques available in the literature and select the most suited technique for CAD models, based on a detailed experimental evaluation.

Development of an off-grid micro-inverter for a scalable photovoltaic kit

Abstract: Taking into account the geographic location of some rural areas, it is not viable, due to technical or economical reasons, to connect certain villages to the electrical grid. Since access to energy is essential for human development, a DC-DC-AC micro-inverter topology for a photovoltaic kit able to address the previous scenario is proposed in this paper. This kit is capable of working in both isolated and parallel modes, with other kits. This solution is also adequate for a post-natural catastrophe scenario, given the power outages associated with these types of phenomena. Each kit consists of a photovoltaic module, an energy storage system and a DC-DC-AC micro-inverter. A control system for all the subsystems is also presented in this paper. Besides that, the management of the DC bus, taking into consideration the photovoltaic system and energy storage system is also proposed. All the theoretical analysis of this kit is detailed in the paper and confirmed through simulations.

Tensor Product-Based Model Transformation Technique Applied to Servo Systems Modeling

Abstract: This paper presents the design and validation of a Tensor Product (TP)-based model of a family of nonlinear servo systems using an appropriate technique. Two parameters of the first principles state-space model of the servo system are optimally tuned using a metaheuristic Grey Wolf Optimizer algorithm in terms of several runs that lead to the parameter intervals. The derivation of the TP model starts with the linear parameter varying model of the servo system, which is next transformed to the strictly speaking TP model, inserted in a series connection with the servo system nonlinearity. The behaviors of the servo system, the TP model and the first principles model are tested in a different scenario to the parameter identification one, and the outputs are measured. The experimental results on a servo system laboratory equipment show that the TP model derived for this system ensures good performance in terms of small relative modeling errors.

Super Twisting Sliding-Mode Control of Grid-Tied Quasi-Z-Source Inverters Under Distorted Grid Voltage

Abstract: A super twisting sliding mode control (STSMC) method is introduced for grid-tied quasi-Z-source inverters (qZSI) operating under distorted grid voltage. This method is capable to alleviate the chattering phenomena in both ac and dc sides of the system concurrently by generating two smooth control inputs for dc and ac sides of the circuit. Applying the STSMC strategy suppresses the chattering to its minimum while preserving the other important features of the sliding mode control method like zero steady-state error, robustness and fast response. Moreover, the ac side capacitor voltage reference is produced by employing a proportional-resonant (PR) controller. An extra term is added into PR controller so as to eliminate the distortions of the grid current under distorted grid voltage. The effectiveness of the proposed method in suppressing the chattering and generating a harmonic free grid current is certified via simulation results using MATLAB/SIMULINK.

Comprehensive Design of a Grid-Connected Seven-Switch Boost-Type Current Source PV Inverter

Abstract: In this paper a grid-connected three-phase seven-switch boost-type photovoltaic (PV) current source inverter (CSI7) is investigated. The designed CSI7 injects sinusoidal currents into the grid with a peak amplitude ${\hat I_g} = 7.71A$ for a maximum PV input-power ${P_{pv - MPP}} = 3.6kW$ at ${V_{pv - MPP}} = 432.48V$ and ${I_{pv - MPP}} = 8.32A$ . The principle of operation is thoroughly described and the voltage and current stresses on the semiconductor power devices are analytically determined. Calculation formula for the design of the passive components are also provided. These calculations are performed based on the inverter's rated operating conditions and their accuracy is verified by digital computer simulations. Accordingly, proper active and passive components are selected and a 3. 6 kW PV-CSI7 is designed. The performance of the designed inverter are evaluated under ideal operating conditions $\left({{T_a} = 25^\circ C,G = 1000W/{m^2}} \right)$ and under shading conditions. An overall efficiency around 98 % is obtained.