The next-generation power system requires converters capable of a dynamic response while meeting multiple control objectives and constraints. These requirements can be systematically handled with model predictive control (MPC), a powerful optimization-based technique. This paper discusses a strategy to co-design an MPC algorithm on heterogeneous compute systems consisting of a processor and a field-programmable array (FPGA). The strategy is to use the processor to execute operations that have a high degree of complexity and data dependency, and the FPGA for operations with low data dependency and possibilities of parallelism. We use high-level synthesis (HLS) to aid the FPGA implementation and show that the strategy is able to synthesize a real-time capable controller for a grid-connected three-phase inverter.

For deploying autonomous mobile robots in the proximity of people, the integration of social norms has been shown to contribute to comfort and socially-aware behavior. In our previous work, we have investigated socially aware collision avoidance based on Dynamical System Modulation (DSM). However, there was a lack of experimental research on how robots can avoid a person in a comfortable manner. In this paper, we extend our previously proposed Socially Aware Dynamical System Modulation (Social-DSM for socially-aware robot navigation by incorporating speed consideration and proactive motion generation based on human intention estimation. The novel framework was implemented on a real robot and validated through a proof-of-concept experiment conducted in a controlled environment, supplemented by a participant survey.

This paper presents a novel framework for socially-aware robot navigation in a pedestrian mixed environment. To improve the acceptance of mobile robots in daily life, robots should exhibit safe and socially acceptable motions to humans, as integrating social norms has been shown to contribute to comfort and socially-aware behavior in various studies. In this paper, we propose a novel framework, Socially-aware Dynamical System based Control (Social-DSC), which leverages the Dynamical System based Control (DSC) for collision avoidance and Social Force Model (SFM) for socially-aware robot navigation. While safety is guaranteed by the DSC framework, the social norms are considered in the modified SFM. In simulation studies, we compared the proposed method with baseline methods, demonstrating higher pedestrian path smoothness and reduced effort in evading motion, thereby improving human comfort.

In this paper, we consider the problem of distributed task allocation for multi robot systems in a manufacturing workshop, where new transportation task may arrive randomly. The tasks of the mobile robot team is to efficiently pick and deliver work pieces according to the production plan. We propose two distributed schemes of receding horizon task allocation (RHTA) algorithm. In our approach, each robot estimates the costs of possible task sequences locally, and resolve task conflicts by assigning priority or constructing a negotiation set based on marginal cost. Monte-Carlo simulations show the performance of the proposed method compared with its centralized counterpart and the popular CBBA method.

Controllers for modern power converters are complex embedded systems. Processor-in-the-loop (PiL) schemes are one way to test and verify the behavior of the controller when executed on the target hardware. PiL schemes can be employed at low cost and can be used to verify if control algorithms are executed correctly and in real-time. Although PiL is often supported by power electronics simulation tools, the range of supported controllers is limited. This paper proposes a framework where PiL schemes can be implemented for a wide range of software tools and controllers. The framework is implemented as a C library with external interfaces that are made configurable and can be adapted to different devices and simulation tools. The C library is open-source and available online. To demonstrate the framework, a PiL scheme is implemented with PLECS as the software tool and a Zynq-7000 system-on-a-chip as the controller.

As the MMC is a highly coupled Multi-Input Multi-Output (MIMO) nonlinear system, classical linear controllers only achieve zero steady-state error and good dynamic performance at a limited linear range around an operation point. Else, its dynamic performance highly depreciates. This paper uses the Input-Output Feedback Linearization (IOFL) technique to decouple the meshed state-dependent inputs of the grid-connected MMC current dynamics model. This method independently controls the DC-, AC-, and Circulating currents with desired performance through the whole operation range. The internal dynamics are analyzed and proven stable. From this analysis, the conduction losses for steady-state operation are obtained as a by-product. Simulation results validate the controller performance under different power and energy exchange scenarios.

Often, the characteristics of thyroid nodules need to be determined by fine needle aspiration (FNA) biopsy. The increasing applications of machine learning and deep learning algorithms provide alternative noninvasive methods to study thyroid nodules on ultrasound images. Many studies examined the feasibility of convolutional neural networks or radiomics feature extraction to analyze the characteristics of thyroid nodules. In this study, we built an automated radiomics analysis system by combining thyroid segmentation via U-Net and radiomics feature extraction. Our proposed machine learning-based automated radiomics analysis was compared to a deep learning-based convolutional neural network method in a two-center thyroid nodule classification task. It is shown that the automated radiomics analysis can accurately segment thyroid nodules to facilitate clinical diagnosis by achieving dice scores of 0.77 and 0.74 on internal and external sets respectively. In addition, the proposed automated radiomics analysis approach can improve sensitivity, negative predictive value (NPV) and positive predictive value (PPV) by 41.2%, 3.5% and 7.5% respectively, while reducing the false negative rate by 41.1%.

Deep learning algorithms can evaluate large and complex sets of data, offering various support for medical imaging analysis. Previous works have explored applications of deep learning to measure leg lengths more efficiently. These previous studies provide evidence to suggest deep-learning algorithms can improve efficiency with high levels of accuracy and speed. In this retrospective study, we utilize deep learning-based convolutional neural networks, programmed with input from a human expert, to identify key points and measure leg length. We collected frontal computed tomography (CT) scout radiographs from pre-operative CT scans of patients undergoing evaluation for knee arthroplasty from diverse sources to both train and test the model. We prepared a DenseNet121 model to predict and identify key points, which were then used to develop patch-based models. We applied separable convolutional layers to complete the analysis. The data reflects that 1) separable convolution exhibits lower mean absolute error (MAE) and increased convergence speed as compared to global average pooling layers and 2) optimal learning rates, batch size, and patch size can be achieved to present the least MAE. Our findings provide useful information and an automated tool to assist radiologists to diagnose leg length discrepancy in clinical practice.

In this paper, we consider the problem of de-centralized control of multi robot systems in the presence of additive bounded uncertainties in the state estimation. First, based on the H ∞ Filter, each robot estimates the state of its neighbors with a guaranteed upper bound. Then, we adopted the Optimal Reciprocal Collision Avoidance (ORCA) to compute a convex collision-avoiding velocity constraint for the local Model Predictive Controller (MPC). The upper bound information of the H ∞ Filter is herein used to modify the ORCA set to ensure a robust collision avoidance. Furthermore, to overcome the influence from estimation error of ego motion as well as local process disturbance, an output feedback tube based MPC scheme is applied, which applies an offline designed auxiliary controller that keeps the actual robot state within a bounded hyper-tube around the nominal trajectory for all times. Finally, simulation results show the effectiveness of the proposed scheme.

A neural radiance field (NeRF) is a scene model supporting high-quality view synthesis, optimized per scene. In this paper, we explore enabling user editing of a category-level NeRF – also known as a conditional radiance field – trained on a shape category. Specifically, we introduce a method for propagating coarse 2D user scribbles to the 3D space, to modify the color or shape of a local region. First, we propose a conditional radiance field that incorporates new modular network components, including a shape branch that is shared across object instances. Observing multiple instances of the same category, our model learns underlying part semantics without any supervision, thereby allowing the propagation of coarse 2D user scribbles to the entire 3D region (e.g., chair seat). Next, we propose a hybrid network update strategy that targets specific network components, which balances efficiency and accuracy. During user interaction, we formulate an optimization problem that both satisfies the user’s constraints and preserves the original object structure. We demonstrate our editing approach on rendered views of three shape datasets and show that it outperforms prior neural editing approaches. Finally, we edit the appearance and shape of a single-view real photograph and show that the edit propagates to extrapolated novel views.

In this paper, the bilateral teleoperation of cooperative manipulators is achieved and experimentally analyzed. The master and slave robots are asymmetrical, and only the master end effector’s task space velocity signals are transmitted through the communication network, while the task space force signals of slave robot are relayed back. A power-based time domain passivity control (PTDPC) approach is employed for the controller design to ensure the passivity of the communication channel in the presence of time-varying delays and are applied to each side of the communication channel at every time constant. This model-free method does not require the dynamic models of the master or slave systems to be known. The slave robot acts as the leader of the remote dual-arm cooperative manipulator system that is used to manipulate a common rigid object. This leader robot is controlled using position control mode to track the trajectory of the master, while the follower robot employs an admittance control method to follow the leader’s motion trend. The follower robot is not required to transmit or receive any communication data, which simplifies the network communication topology. Experimental results are presented to verify the effectiveness and simplicity of the designed framework in the presence of large, time-varying and asymmetric delays.

App builders commonly use security challenges, a form of step-up authentication, to add security to their apps. However, the ethical implications of this type of architecture has not been studied previously.In this paper, we present a large-scale measurement study of running an existing anti-fraud security challenge, Boxer, in real apps running on mobile devices. We find that although Boxer does work well overall, it is unable to scan effectively on devices that run its machine learning models at less than one frame per second (FPS), blocking users who use inexpensive devices.With the insights from our study, we design Daredevil, a new anti-fraud system for scanning payment cards that works well across the broad range of performance characteristics and hardware configurations found on modern mobile devices. Daredevil reduces the number of devices that run at less than one FPS by an order of magnitude compared to Boxer, providing a more equitable system for fighting fraud.In total, we collect data from 5,085,444 real devices spread across 496 real apps running production software and interacting with real users.

This article proposes a targeting posture control approach with dynamic obstacle avoidance of differential-drive wheeled mobile robot (WMR) systems in the presence of unknown skidding, slipping, input disturbances, model uncertainties, and torque saturation. First, a nonlinear model predictive control (NMPC) scheme is presented to generate a feasible trajectory from a starting posture to a targeting posture where dynamic obstacles in the environment and physical constraints of the robots are considered. Second, a robust virtual control law at the kinematic level is introduced for the robots to follow the trajectory. Third, taking the consideration of the unknown skidding, slipping, input disturbances, and model uncertainties in the dynamic model being lumped as a total disturbance, which is estimated by the linear extended state observer (LESO), a disturbance compensation-based saturation controller is designed to make the real velocity of the robot converge to the virtual velocity command. Finally, the effectiveness of the proposed control strategy is verified by simulation results.

We introduce a simple but effective unsupervised method for generating diverse images. We train a class-conditional GAN model without using manually annotated class labels. Instead, our model is conditional on labels automatically derived from clustering in the discriminator’s feature space. Our clustering step automatically discovers diverse modes, and explicitly requires the generator to cover them. Experiments on standard mode collapse benchmarks show that our method outperforms several competing methods when addressing mode collapse. Our method also performs well on large-scale datasets such as ImageNet and Places365, improving both diversity and standard metrics (e.g., Fréchet Inception Distance), compared to previous methods.

Trajectory planning for an autonomous bicycle is investigated with regard to different vehicle models. To this end, a nonlinear optimization problem is set up using three models which differ with respect to the geometric details and, therefor, the included physical effects. First, using the method of Direct Collocation the optimization problem is transcribed with the vehicle models as equality constraints. The resulting trajectories are applied to the most detailed vehicle model using a 2-degree-of-freedom loop structure for validation and comparison. Further, a prototype autonomous two-wheeled vehicle is introduced which was developed for experimental verification of motion planning and control algorithms. Finally, experiments are presented and discussed that are run on the real vehicle for a particular maneuver. Thereby, the differences between the trajectories created by different vehicle models are emphasized.

As key factors to guarantee accurate localization for ultra-wide band system (UWB), Non-line-of-sight (NLOS) identification and mitigation attract lots of attentions. One of the most effective methods for NLOS detection is based on the different characters of channel impulse response (CIR) under Line-of-sight (LOS) and NLOS condition. Features (such as kurtosis, standard deviation, energy, etc.) extracted from CIR are used for classification with the help of machine learning algorithm. Different from existing approaches, the NLOS and LOS probability density functions (PDF) of the correlation coefficient are calculated with the training data. The probability that the CIR is measured under LOS or NLOS is determined based on the PDF. A weighted particle filter is proposed to reduce the localization error, caused by NLOS. The weights for the proposed approach are obtained with the help of the measured variance of the LOS error and the NLOS/LOS probability. The weighted least squares (WLS) and standard particle filter are used as comparison. The real field test results show that the proposed weighted particle filter has better accuracy compared to the standard particle filter and the WLS.

Model order reduction is a very useful technique for obtaining reduced sized models that are mathematically equivalent to the original one, with a reasonably small error. However this reduction is done in a generalized manner, without considering the purpose of the application. Reduced models obtained in such a way, in most cases, are not accurate enough for fault isolation. This paper presents a method that takes the faults into consideration during the model order reduction process in an optimal way, allowing proper fault detection and isolation to be achieved.

In this paper, a set-up consisting of a medium voltage grid with connected water distribution network as a load is investigated. The water distribution network contains water reservoirs that can be used as storages to support the voltage control by adjusting the electric power consumption to keep the voltages in the required limitations. At the same time the electric power and water demand are satisfied. The considered set-up is derived from a real-world set-up in the city of Kaiserslautern, Germany, where the power grid and water distribution grid are operated by the same operator. A distributed model predictive controller for the voltage control is designed in general form and thus the concept be applied to other loads containing storages. The controller is tested in simulations.

In order to handle the challenges associated with large capacities wind farms integrated in the power grid, different techniques such as optimization algorithms, artificial intelligence and others are employed and studied. Recently, many researches focus on the frequency regulation problem by showing the role of kinetic energy stored in the rotor side of wind turbine in supporting the grid frequency. The main challenge of the inertia emulation scheme is to release the maximum possible amount of kinetic energy without violating operating limits of the wind turbines and keeping in mind the grid integration standards for wind farms. However, no selection criteria for the control parameters has yet been given. The objective in this paper is to consider the constraints and limitations on the rotor speed and buffered power in DFIG wind turbines and also to highlight the influence of the controller parameters on the stability of the power grid and the operation of the wind farms. The stability of the power system with high penetration of wind farms is analyzed based on Kuramoto theory and a sufficient stability condition is derived and mathematically proved. Finally the stability condition is verified in a test power system with a large integration of wind power.

This paper proposes a passivity-based trajectory tracking control for an autonomous bicycle. The nonlinear bicylce model is considered as an under-actuated pseudo-Hamiltonian system including the nonholonomic constraints. For the synthesis of the tracking controller the method of generalized canonical transformation is applied with some modifications. A first transformation compensates the effect of the gravity and, then, creates an error system. This reduces the problem of tracking to a stabilisation problem. A second transformation stabilises the resulting time-varying error system on the trajectory. Further, an integral action is added to deal with uncertainties and disturbance. Finally, simulations are run for two types of trajectories to verify the stability and investigate the performance of the closed loop.

The location estimation of a mobile station (MS) is becoming more and more important in many applications. By using the time difference of arrival (TDOA) method, the position of a MS can be determined based on the relative time difference measurements. However, the accuracy of the position estimation suffers from Non-line-of-sight (NLOS) errors. In this paper, the influence of the main base station (BS) selection for the position estimation accuracy is discussed. It is shown that the selection of the accurate range differences is more important for position estimation than the detection of the NLOS BSs for TDOA method. A novel accurate range differences selection approach based on IMU measurements is developed. Instead of selecting the BSs under line-of-sight (LOS), the accurate range differences are chosen for calculation. The selection method is combined with Iterative Extended Kalman Filter (IEKF) to effectively mitigate the NLOS measurement errors. The IEKF performance difference, with and without the selection approach, is analyzed. The performance of ultra-wideband/inertial sensor (UWB/INS) tightly coupled approach and Taylor approach are discussed. Among all the approaches, the IEKF with selection approach shows better accuracy both in simulation and real field tests.

This paper presents the application of a modular multilevel cascade converter(MMCC) based on single-delta bridge cells(SDBC) as an active power filter(APF). It is required that the mean value of the DC capacitor voltages is controlled so that the tracking of the source current reference is guaranteed in operation of APF. This paper proposes a DC capacitor voltage regulation method for MMCC-SDBC under different PCC voltages, including the ideal, distorted and unbalanced situations. With the application of the instantaneous symmetrical component theory, this paper analyzes the currents and average active power in three branches of MMCC-SDBC. Based on the analysis and the energy-power relationship of capacitors, the paper develops a systematic regulation method to maintain the DC-bus voltage on a fixed level. The simulation in MATLAB of a three-phase system under different PCC voltage conditions feeding a nonlinear load has shown the effectiveness of the proposed method.

The aperiodic optimal linear estimator design problem is investigated in this paper for networked systems with communication uncertainties, including delays and data losses, where the sampling and estimation are nonuniform and asynchronous. Based on the idea of measurement fusion, two approaches are proposed to design the aperiodic estimators, and it is shown that the estimator is equivalent to that designed by the measurement augmentation method in performance. Moreover, the estimation performance is improved by using a newly proposed measurement retransmission scheme as compared with the commonly used hold-input and zero-input schemes, by which the lost measurements are never used once they are lost.

The convex hull of a set of two-dimensional points, P, is the minimal convex polygon that contains all the points in P. Convex hull is important in many applications such as GIS, statistical analysis and data mining. Due to the ubiquity of data uncertainty such as location uncertainty in real-world applications, we study the concept of convex hull over uncertain data in 2D space. We propose the Probabilistic Convex Hull(PCH) query and demonstrate its applications, such as Flickr landscape photo extraction and activity region visualization, where location uncertainty is incurred by GPS devices or sensors. To tackle the problem of possible world explosion, we develop an O(N3) algorithm based on geometric properties, where N is the data size. We further improve this algorithm with spatial indices and effective pruning techniques, which prune the majority of data instances. To achieve better time complexity, we propose another O(N2log N) algorithm, by maintaining a probability oracle in the form of a circular array with nice properties. Finally, to support applications that require fast response, we develop a Gibbs-sampling-based approximation algorithm which efficiently finds the PCH with high accuracy. Extensive experiments are conducted to verify the efficiency of our algorithms for answering PCH queries.

In this paper, a two-degree-of-freedom-control concept for a flexible single-link robot is considered offering trajectory tracking as well as vibration damping. The robot - a flexible link with tip mass attached to a moving cart - is modeled as distributed parameter system including structural damping. The proposed forward controller is designed based on the exact modal approximation of the system dynamics. Based on the infinite dimensional error dynamics formulated in strict feedback form, a backstepping controller is designed that guarantees passivity. Input Shaping is used for planning the desired trajectories. Measurement results at a laboratory experiment show the feasibility of this approach.

STATic synchronous COMpensator (STATCOM) can be integrated into electric transmission systems to provide reactive power compensation and grid voltage support for achieving high-efficient and reliable Flexible AC Transmission Systems (FACTS). An emerging solution for STATCOM with modular structure and multilevel voltage output, named as modular multilevel STATCOM (mmSTATCOM) shows its great advantages of adaption to a wide voltage range, sinusoidal output voltages with less harmonics and less converter losses, compared with the traditional STATCOM. The cascaded Full-Bridge-based mmSTATCOM that is configured with three converter branches into a single delta connection can be classified into the modular multilevel cascade converter (MMCC) family and this configuration can be named as Single-Delta Full-Bridge (SDFB). This paper presents a novel integrated modeling method of SDFB-based STATCOM by taking the independent currents, the total energy and the internal energy balancing into consideration. Then a nonlinear multivariable state-space model is developed. Therefore, a nonlinear control method, named as nonlinear quadratic regulator (NQR), is accordingly proposed to guarantee a safe long-term operation of the SDFB-based STATCOM system both in the balanced and temporarily unbalanced grid conditions. The simulation results are followed to verify the system operations in the normal condition as well as during a temporary grid fault.

Decentralized control approaches arise as promising solutions to cope with the high complexity of interconnected networks such as in the traffic domain. Leaving the control of traffic flow in the hands of distributed traffic nodes opens up the potential for a more efficient network utilization and a higher adaptability to changing traffic conditions. However, some control approaches based on decentralized information handling will introduce information delays which can result in suboptimal system performance and unintended effects such as oscillating traffic flows and traffic congestions. This paper compares the performance of a robust PI control approach without and with Smith predictor for an automated traffic control node with delayed traffic information. Results are validated with a realistic traffic simulation tool.

Given two sets of points, S and T, in a road network, G, a distance-preserving subgraph (DPS) query returns a subgraph of G that preserves the shortest path from any point in S to any point in T. DPS queries are important in many real world applications, such as route recommendation systems, logistics planning, and all kinds of shortest-path-related applications that run on resource-limited mobile devices. In this paper, we study efficient algorithms for processing DPS queries in large road networks. Four algorithms are proposed with different tradeoffs in terms of DPS quality and query processing time, and the best one is a graph-partitioning based index, called RoadPart, that finds a high quality DPS with short response time. Extensive experiments on large road networks demonstrate the merits of our algorithms, and verify the efficiency of RoadPart for finding a high-quality DPS.

Modular multilevel converter (MMC) is considered as an idea and promising candidate for multilevel technology. By connecting two MMCs in back-to-back configuration with a DC transmission cable, a so-called indirect MMC can be constructed for high-voltage direct current transmission application. This paper presents the dq model of indirect MMC by generalized state-space representation. Correspondingly, a complete LQR-based state feedback control theory is applied to regulate the multiple currents in indirect MMC. The resulted control architecture is unified and flexible to include other control criterion while guaranteeing an optimized control performance. The grid-side unbalanced situation of indirect MMC is analyzed by mathematical modeling and an estimator is proposed to reconstruct the unbalanced disturbances. Simulation results are presented to illustrate the effectiveness of estimator-based LQR controller under both balanced and unbalanced conditions.

This paper presents an energy management system for smart grids with electric vehicles based on hierarchical model predictive control (HiMPC). The energy management system realizes load-frequency control (LFC), an economic operation and an electric vehicle integration into the smart grid. The main component is the HiMPC, which allows covering different time scales, regarding constraints (e.g. power ratings) and predictions (e.g. on renewable generation), as well as rejecting disturbances (e.g. due to fluctuating renewable generation) based on a systematic model- and optimization-based design. For the electric vehicle integration, an aggregator is proposed as link between HiMPC and individual vehicle. The aggregator in particular provides predictions to the HiMPC on the availability of electric vehicles for LFC based on the current mobility demand and the statistical mobility behavior of the vehicle users. Throughout the paper, the energy management system is evaluated for the smart grid of an intermediate city.

An approach for a leak detection and localization scheme is presented. With respect to developing methods for diagnosis in commercial vehicles, applicability of existing approaches for leak diagnosis in pipelines has been evaluated. A leakage localization strategy using an Extended Kalman Filter (EKF) in combination with the Method of Characteristics (MOC) is adapted to pneumatic systems and implemented in Matlab/Simulink. To deal with boundary limitations due to possible sections for leak localization along the pipe, a methodical extension based on the idea of virtual pipes is proposed and integrated into the overall method. Verifications, resulting from a simulation model as well as from measurements at a real test bench, are presented. The promising results are discussed and suggestions for possible steps of further development are given.

The paper proposes a hybrid Kalman filter integrating a robust and optimal algorithm for the use as an observer in model-varying predictive control (MVPC) of a nonlinear system. Moreover, a position MPC is derived in detail. Even though the proposed approach is quite general, a real case coming from automotive application is studied using computer simulation to demonstrate the effectiveness of the proposed technique. Simulations and results with real data are also discussed.

The paper presents a novel technique to control the current of an electromagnetic linear actuator fed by a multilevel IGBT voltage inverter with dynamic energy storage. In order to provide short response time, high precision and low switching frequency at the same time we combine current hysteresis regulation with model predictive control (MPC). The proposed MPC technique works in very short receding horizon and includes also wavelet algorithm to optimize the MPC operation. Through simulations with real actuator data the proposed technique shows very promising results.