Feedforward Method Research Articles

Application of machine learning approach (artificial neural network) and shrinking core model in cobalt (II) and copper (II) leaching process

The leaching laboratory experiment uses the artificial neural network (ANN) to predict and evaluate copper and cobalt recovery. This study aimed to evaluate the efficacy of using the shrinking core model in conjunction with an artificial neural network (ANN) as part of a machine learning strategy to improve the leaching process of cobalt (II) and copper (II). The numerous factors in the leaching process, such as acid concentration, leaching time, temperature, soil-to-solution ratio, and stirring speed, are adjusted using an ANN with several layers, feed-forward, and back-propagation learning methods. These variables are in charge of the high cobalt recovery during the reduced sulfuric acid leaching procedure. The ANN algorithm has 10 hidden layers, 5 input variables describing the leaching parameters, and two neurons as output layers corresponding to copper and cobalt leaching recovery. The optimum conditions were found to be acid concentration of 100 g/L, leaching duration 120 min, temperature 55 °C, soil-to-solution ratio of 1:40 g/mL, and stirring speed 300 rpm. The optimized trained neural networks tested, trained, and validated steps are represented by R 2 values of 0.94, 0.99, 0.97, and 0.97, respectively, equating to 97.5% copper recovery and 95.4% cobalt recovery.

Feed-forward networks using logistic regression and support vector machine for whole-slide breast cancer histopathology image classification

The performance of an image classification depends on the efficiency of the feature learning process. This process is a challenging task that traditionally requires prior knowledge from domain experts. Recently, representation learning was introduced to extract features directly from the raw images without any prior knowledge. Deep learning using a Convolutional Neural Network (CNN) has gained massive attention for performing image classification, as it achieves remarkable accuracy that sometimes exceeds human performance. But this type of network learns features by using a back-propagation approach. This approach requires a huge amount of training data and suffers from the vanishing gradient problem that deteriorates the feature learning. The forward-propagation approach uses predefined filters or filters learned outside the model and applied in a feed-forward manner. This approach is proven to achieve good results with small size labeled datasets. In this work, we investigate the suitability of using two feed-forward methods such as Convolutional Logistic Regression Network (CLR), and Convolutional Support Vector Machine Network for Histopathology Images (CSVM-H). The experiments we have conducted on two small breast cancer datasets (Sultan Qaboos University Hospital (SQUH) and BreaKHis dataset) demonstrate the advantage of using feed-forward approaches over the traditional back-propagation ones. On those datasets, the proposed models CLR and CSVM-H were faster to train and achieved better classification performance than the traditional back-propagation methods (VggNet-16 and ResNet-50) on the SQUH dataset. Importantly, our proposed approach CLR and CSVM-H efficiently learn representations from small amounts of breast cancer whole-slide images and achieve an AUC of 0.83 and 0.84, respectively, on the SQUH dataset. Moreover, the proposed models reduce memory footprint in the classification of Whole-Slide histopathology images since their training time is significantly reduced compared to the traditional CNN on the SQUH and BreaKHis datasets.

A Novel Friction Compensation Method Based on Stribeck Model With Fuzzy Filter for PMSM Servo Systems

Friction deteriorates speed and position control accuracy of high-precision machine tools and causes distortion phenomenon when the speed crosses zero. In this paper, a new compensation method based on static Stribeck model (SB) combined with fuzzy low-pass filter (SBFF) is studied for friction compensation of permanent magnet synchronous motor (PMSM) servo systems. First, models of the servo system, the Stribeck, and the LuGre are established. Then, it is theoretically derived that the hysteresis effect of the LuGre model (LG) can be equivalent to a low-pass filter term compared to the SB. Based on this, the relationship between the filter time constant and the motor reference speed can be derived. To further improve the friction compensation effect near zero speed and to weaken the delay generated by the filter during the non-zero-speed range, fuzzy logic is introduced to adjust the filter time constant according to the reference speed variation. Through simulations and experiments, compensation strategies based on the LG feedforward, SB feedforward, SBFF feedforward and SBFF feedback are compared, and the robustness of the proposed SBFF feedforward compensation method over entire speed range is verified as well.

Research on Friction Compensation Method of Electromechanical Actuator Based on Improved Active Disturbance Rejection Control

The friction factor of harmonic reducers affects the transmission accuracy in electromechanical actuators (EMAs). In this study, we proposed a friction feedforward compensation method based on improved active disturbance rejection control (IADRC). A mathematical model of EMA was developed. The relationship between friction torque and torque current was derived. Furthermore, the compound ADRC control method of second-order speed loop and position loop was studied, and an IADRC control method was proposed. A real EMA was developed, and the working principles of the EMA driving circuit and current sampling were analyzed. The three methods—PI, ADRC, and IADRC—were verified by conducting speed step experiments and sinusoidal tracking experiments. The integral values of time multiplied by the absolute error of the three control modes under the step speed mode were approximately 47.7, 32.1, and 15.5, respectively. Disregarding the inertia of the reducer and assuming that the torque during no-load operation equals the friction torque during constant motion, the findings indicate that, under a load purely driven by inertia, the IADRC control method enhances tracking accuracy.

Non-parametric style transfer: Correlation-aware exact distribution matching

Early feed-forward neural methods of arbitrary image style transfer utilized the encoded feature map up to its second-order statistics, i.e., mean and variance (or covariance) of the encoded feature map. Recent methods have begun to utilize feature statistics beyond the 2nd-order statistics, e.g., a higher-order moment, slicing-based method, and histogram matching for better style representation. However, even in the distribution matching of the latest method, considering the channel-correlation and the order of duplicated values in the feature map is missing. This resulted in insufficient style transfer or artifacts on style-transferred images e.g., unexpected color or pattern co-occurrence, skewed color or style pattern into one side, noise between background and foreground, etc. In this work, we show how to make a feature transform layer to consider both channel-correlation and duplicated values in the feature map for exact feature distribution matching in style transfer. Our experimental results prove that the stylized images obtained with our method are more similar to the target style images qualitatively without losing content clearness and superior to the existing methods quantitatively in all possible style measures, e.g., 0.0311 lower sliced Wasserstein distance than the previous method and the most favored choice with 25.54% preference.

Graph neural network-based similarity relationship construction model for geospatial services

ABSTRACT During the development of service-based software systems, Geospatial Service (GS) replacement is often performed, which requires the discovery of functionally similar services in service registries to replace failed services. Compared to real-time similarity computations, direct extraction of similar services from constructed similarity relationships can yield higher replacement efficiency. However, missing and inconsistent service-registry information impedes accurate similarity relationship construction. Here, we propose a Graph Neural Network (GNN)-based model for GS Similarity Relationship construction considering service descriptions and tags, which is named GSSR-GNN. As the sparsity of the service similarity relationship graph constructed based on labeled samples limits the information propagation ability, a graph augmentation method for similarity relationship construction among second-order neighbors is proposed. Considering the differences in the semantic-information feature distributions, such as the service descriptions and tags, a feed-forward neural network-based fusion method is designed to embed them into the same vector space. Pre-trained Bidirectional Encoder Representations from Transformers (BERT) and WordNet models are introduced to enhance the service-representation expressiveness. When an enhanced service representation is input to the GNN, the similarity is calculated and the service similarity relationship is obtained. Experimental results show that the proposed model constructs service similarity relationships with high precision, thus improving the service replacement efficiency and reducing the computational cost of service registry during service replacement.

Analysis of torque controlling strategies of interior permanent magnet synchronous machine in hybrid electric vehicle

Significance of this research is based on torque controlling strategies in hybrid electric vehicle. The traction motor of a hybrid electric commercial vehicle must have a high power density-to-weight ratio and excellent torque at extremely low speeds. A proportional-integral controller estimates the high-resolution rotor position based on the difference between actual as well as reference output power. This technique can correct for hall position sensor installation error and also accurately estimate rotor position. In this study, a current proportional integral controller is also used to determine the permanent magnetic flux linkage. Experiments are used to test other key metrics such as d-axis and q-axis inductances, stator resistance, and energy loss. The measured parameters are kept in lookup tables that span the whole operating range at various current levels. A maximum torque per ampere control methodology, paired with the feed forward parameter iteration method, may be used to produce accurate and efficient torque control based on these precise parameters. The standard driving cycle performance requirements of the modelled Federal Test Procedure vehicle is analysed in the article for understanding the real model requirements of the vehicle. The efficacy of the proposed methodology using Matlab/Simulink environment is analysed. Both modelling and experimental findings validate the efficacy, torque ripple, stator and rotor flux, sensitivity of the vehicle parameter as well as the computational complexity of the proposed methodology.

Your smartphone could act as a pulse-oximeter and as a single-lead ECG

In the post-covid19 era, every new wave of the pandemic causes an increased concern/interest among the masses to learn more about their state of well-being. Therefore, it is the need of the hour to come up with ubiquitous, low-cost, non-invasive tools for rapid and continuous monitoring of body vitals that reflect the status of one’s overall health. In this backdrop, this work proposes a deep learning approach to turn a smartphone—the popular hand-held personal gadget—into a diagnostic tool to measure/monitor the three most important body vitals, i.e., pulse rate (PR), blood oxygen saturation level (aka SpO2), and respiratory rate (RR). Furthermore, we propose another method that could extract a single-lead electrocardiograph (ECG) of the subject. The proposed methods include the following core steps: subject records a small video of his/her fingertip by placing his/her finger on the rear camera of the smartphone, and the recorded video is pre-processed to extract the filtered and/or detrended video-photoplethysmography (vPPG) signal, which is then fed to custom-built convolutional neural networks (CNN), which eventually spit-out the vitals (PR, SpO2, and RR) as well as a single-lead ECG of the subject. To be precise, the contribution of this paper is twofold: (1) estimation of the three body vitals (PR, SpO2, RR) from the vPPG data using custom-built CNNs, vision transformer, and most importantly by CLIP model (a popular image-caption-generator model); (2) a novel discrete cosine transform+feedforward neural network-based method that translates the recorded video-PPG signal to a single-lead ECG signal. The significance of this work is twofold: (i) it allows rapid self-testing of body vitals (e.g., self-monitoring for covid19 symptoms), (ii) it enables rapid self-acquisition of a single-lead ECG, and thus allows early detection of atrial fibrillation (abormal heart beat or arrhythmia), which in turn could enable early intervention in response to a range of cardiovascular diseases, and could help save many precious lives. Our work could help reduce the burden on healthcare facilities and could lead to reduction in health insurance costs.

A Nonlinear Load Current Feedforward Strategy for the Charge-Controlled LLC Converter and Its Digital Implementation to Improve the Dynamic Response

Compared with the traditional single voltage loop control strategy in LLC converters, the charge-controlled LLC converter can significantly improve its dynamic response due to its model having the first-order feature. Generally, a load current feedforward method could improve the converter's dynamic response. However, for the LLC converter, it is almost impractical due to the unclear relationship between the load current and the control variable. To build this relationship, a mathematical expression between the load current change and the control variable is derived that can precisely offset the load disturbance and achieve faster dynamic response. The derived relationship is nonlinear, related to both input and output voltages and switching frequency, and is suitable for digital implementation. To further analyze the dynamic response, the output impedances with and without the load current feedforward control are established and analyzed, which visually demonstrates the dynamic improvement in the time domain through the calculation of the output impedance and load change spectrum in the frequency domain. Finally, a 120W LLC prototype with nominal 48V input and 12V output is built to verify the nonlinear load current feedforward control and its analysis.

An 8-Channel Ambulatory EEG Recording IC With In-Channel Fully-Analog Real-Time Motion Artifact Extraction and Removal.

Wereport the design, implementation, and experimental characterization of an 8-channel EEG recording IC (0.13 μm CMOS, 12 mm 2 total area) with a channel architecture that conducts both the extraction and removal of motion artifacts on-chip and in-channel. The proposed dual-path feed-forward method for artifact extraction and removal is implemented in the analog domain, hence is needless of a DSP unit for artifact estimation, and its associated high-DR ADCs and DACs employed by the state of the art for artifact replica generation. Additionally, the presented architecture improves system's scalability as it enables channels' stand-alone operation, and yields the lowest reported channel power consumption among works featuring motion artifact detection/removal. Following an experimental study on electrode-skin interface electrical characteristics for dry electrodes in the absence and presence of motions, the article presents the channel architecture, its detailed signal transfer function analysis, circuit-level implementation, and experimental characterization results. Our measurement results show an amplification voltage gain of 48.3 dB, a bandwidth of 300 Hz, rail-to-rail input DC offset tolerance, and 41.5 dB artifact suppression, while consuming 55 μW per channel. The system's efficacy in EEG motion artifact suppression is validated experimentally, and system- and circuit-level features and performance metrics of the presented design are compared with the state of the art.

A calibration method for predicting and compensating the flexible positioning errors of a 7-DOF series compliant robot caused by end-effector’s gravity

Due to low stiffness of compliant series robot body, the influence of joint flexible deformation caused by the gravity load of the end-effector on the absolute positioning accuracy cannot be ignored. In order to improve the flexible positioning accuracy of the robot end-effector, a calibration method for predicting and compensating the flexible positioning error of the 7-DOF series compliant robot due to the gravity load of the end-effector is proposed. The method identifies and predicts the pose configuration with similar flexible positioning error through the semi-parametric error model based on joint stiffness identification and of BP neural network. The feedforward method is used to compensate the positioning error. In this study, a 7-DOF series compliant robot is selected as the research object. Based on the principle of error similarity, the measurement points and the corresponding pose configurations are selected as the training data for the parameter identification of the semi-parametric error model, and the flexible positioning errors on the continuous trajectory of the end-effector are compensated according to the predicted values of the model. When the mass of the end-effector is 5 kg, by this method, the maximum value, mean value and mean square error of flexible positioning error on multiple straight-line trajectories on the plane are increased from 8.4779, 5.6039, and 1.8496 mm to 0.9720, 0.3079, and 0.2303 mm, respectively. It means the positioning accuracy are increased by 88.53%, 94.50%, and 87.55%, respectively, by which could observe that the precision of this robot’s end-effector was significantly improved.

An adaptive feedforward method in active vibration control for the propulsion shafting system

This paper proposes a new adaptive feedforward control method for the rejection of multiple periodic disturbances in the propulsion shafting system. The active vibration control scheme of the propulsion shafting system is established by employing the equivalent-input-disturbance principle which reduces the complexity of the control system in applications. The adaptive control algorithm is designed by taking the scale transformation procedure and the frequency-tracking technique into consideration. The scale transformation normalizes the iso-surface of the optimization function by the frequency response characteristics of the control channel, which immensely improves the convergence rate of the control algorithm for multi-harmonic disturbance. Moreover, the adaptive controller tracks the fluctuating frequencies in a direct method, enhancing the efficiency and robustness of the algorithm for eliminating time-varying disturbances. Finally, numerical simulation and experimental validation are carried out to show that the proposed method has good performance on the application of the active vibration control of the propulsion shafting system.

Experimental validation of passive and active fault-tolerant controls against sensor faults in a proton exchange membrane fuel cell system

A fuel-cell system is interconnected with subsystems such as the stack, thermal/water management systems, air/fuel supply systems, and inverter; if an abnormality occurs, operational efficiency and availability of the entire system can decrease. However, when faults occur in the system, terminating system operations or replacing faulty parts every time can be expensive and inefficient. Instead, appropriate decision-making is required, with fault-tolerant control considered an appropriate solution. This study presents methods and criteria for fault-tolerant control when a fault occurs in a proton exchange membrane fuel cell (PEMFC) system. By configuring a 1 kW residential PEMFC system, a fault occurring in either the thermocouple or mass flow meters – targets for the system control – is diagnosed, and real-time control change is performed instead of immediately repairing the sensor. This approach enables stable and continuous operation of the PEMFC system. The fault-tolerant control method consists of either the feed-back control method, using the real-time predicted value of each control sensor, or the feed-forward method, using the constant control-signal value learned in the normal state; the advantages and limitations of each method are compared. Futhermore, considering the system efficiency, stack durability, and operating control deterioration, a guideline is suggested for applying fault-tolerant control and improving its corresponding effectiveness. For example, fault-tolerant control can resolve a 4.6% decrease in system efficiency caused by a thermocouple fault. This study contributes to improving the system operation rate and reliability by applying fault-tolerant control, thus facilitating efficient decision-making following fault diagnosis.

Grid Harmonics Suppression for Three Phase Dual-Frequency Grid-Connected Inverter Based on Feedforward Compensation

Using a low pulse ratio, the electromagnetic interference and switching loss of an inverter can be effectively reduced, particularly in high-power applications. However, due to variations in grid impedance, it is a challenging task to achieve stable operation of an LCL-type grid-connected inverter (GCI) using the active damping method with low pulse ratio. Thus, a novel three-phase dual-frequency GCI is presented to ensure the symmetry of the output power and the stable operation of the system address stability issues. The proposed inverter topology in this article is composed of two inverters in parallel, which are, respectively, a power inverter unit (PIU) and an auxiliary harmonic elimination unit (AHEU). To reduce the switching loss and improve the inverter efficiency, the switching frequency of the PIU is relatively low, injecting current into the grid. Moreover, the feedforward compensation method is used in AHEU. AHEU operates at high switching frequency to generate a current component that is symmetrical with the ripple com-ponent, improving the power quality, without extracting the current harmonic as the current reference. The operating principle of feedforward compensation is explained, and a proper parameter design procedure is presented in this paper. Since L filters are used for the proposed inverter, the system can operate stably where the ratio of switching frequency to fundamental frequency is low. A 10 kW laboratory prototype was built. The experimental results showed that the grid current ripple could be effectively eliminated and the THD of the grid current was 3.01%. The proposed inverter has good stability in a weak grid, and the efficiency of the proposed inverter is 95.98% at rated current, which is 0.81% higher than the traditional GCI, effectively increasing the efficiency of the system.

Finite-Time Output Regulation of Linear Heterogeneous Multi-Agent Systems Using Smooth Control

In this paper, we investigate a finite-time output regulation (FTOR) problem for linear heterogeneous multi-agent systems (LHMAS). The objective is to design protocols such that each agent can follow the reference signal within a finite time. An exosystem is given to generate the reference signal and disturbances. Only a part of agents are assumed to have access to the system matrix and state of the exosystem. First, smooth distributed finite-time observers are designed to estimate them. Then, with the help of feedforward method, smooth state feedback and output feedback controllers are designed to address the FTOR problem. By virtue of the finite-time control method and the OR technique, it is shown that the proposed smooth distributed observers and controllers can achieve the finite-time convergence of the estimation errors and error outputs, respectively. Finally, a numerical example is presented for illustration.

A Generalized and Efficient Control-Oriented Modeling Approach for Vibration-Prone Delta 3D Printers Using Receptance Coupling

Delta 3D printers can significantly increase throughput in additive manufacturing by enabling faster and more precise motion compared to conventional serial-axis 3D printers. Further improvements in motion speed and part quality can be realized through model-based feedforward vibration control, as demonstrated on serial-axis 3D printers. However, delta machines have not benefited from model-based controllers because of the difficulty in accurately modeling their position-dependent, coupled nonlinear dynamics. In this paper, we propose an efficient framework to obtain accurate linear parameter-varying models of delta 3D printers at any position within their workspace from a few frequency response measurements. We decompose the dynamics into two sub-models&#x2013;(1) an experimentally-identified sub-model containing decoupled vibration dynamics; and (2) an analytically-derived sub-model containing coupled dynamics&#x2013;which are combined into one using receptance coupling. We generalize the framework by extending the analytical model of (2) to account for differing mass profiles and dynamic models of the printer&#x2019;s end-effector. Experiments demonstrate reasonably accurate predictions of the position-dependent dynamics of a commercial delta printer, augmented with a direct drive extruder, at various positions in its workspace. <i>Note to Practitioners</i>&#x2014;This work aims to equip high-speed 3D printers, like delta machines, with model-based controllers to complement their speed with high-accuracy. Due to the coupled kinematic chains of the delta, complex control methodologies, some of which require real-time state measurements, are often used to achieve satisfactory control performance. Our modeling approach provides an efficient methodology for obtaining accurate linear models without the need for real-time measurements, thus enabling practitioners to design linear model-based feedforward controllers to achieve the high throughput and accuracy desired in additive manufacturing (AM). The models we develop in this paper are intended for use with feedforward vibration compensation methods, which can be beneficial for both industrial-scale AM machines that have high-powered servo motors and feedback controllers, as well as consumer-grade AM machines which use stepper motors in feedforward control.

A Parallel Solution of Timing Synchronization in High-Speed Remote Sensing Data Transmission

Considering the problem that the timing synchronization calculation in high-speed remote sensing signal reception is complex and it is difficult for it to be parallel, this paper deduces and designs a parallel timing error estimation and correction scheme. This paper presents the design of polyphase DFT filter banks with non-maximum decimation. The feedforward timing error estimation and correction method is then improved to enhance synchronization performance. Finally, an implementation scheme for parallel timing error estimation and correction is proposed using the polyphase filter bank time domain decomposition method and the filter polyphase model. In the estimation module, the parallel implementation structure of the joint second-order and fourth-order cyclic statistics methods is designed, which improves the estimation accuracy. In the correction module, a fractional delay filtering method with higher accuracy is adopted in order to improve the calibration accuracy and reduce the computational complexity. The timing synchronization of a high-speed remote sensing signal with timing error is simulated and verified. The experimental results show that the parallel method proposed in this paper greatly reduces the processing speed of subband data, and has a good synchronization performance, which is close to the theoretical limit in the demodulation error rate. This paper utilizes a multi-phase DFT filter bank architecture to achieve parallel timing synchronization, which presents a novel approach for the future parallel reception of high-speed remote sensing signals.

Klasifikasi Ujaran Kebencian Menggunakan Metode FeedForward Neural Network (IndoBERT)

Everyone in Indonesia has freedom of speech, both in real life and on social media. However, freedom of speech carried out without filtering can lead to hate speech. Hate speech is a form of discrimination directed against individuals or groups of individuals based on race, religion, gender, sexual orientation, or other identities. Hate speech can harm other parties which as a result can trigger conflict, violence, and can even cost a person's life. Therefore, it is important to be able to identify and manage this hate speech effectively. One way to manage hate speech on social media is to classify it. In this study, a web-based application was created that can classify a sentence to determine whether the sentence is hate speech or a normal sentence. The model created for classification uses the feedforward neural network method with IndoBERT. Based on the test results, the model created using the feedforward neural network method with IndoBERT provides the best accuracy of 89.52%.

A novel whale optimization smith predictor based on feedforward compensation for controlling wave rotor refrigeration process

AbstractAiming at the difficulty of temperature control and disturbance rejection control in wave rotor refrigeration process (WRRP), a novel whale optimization two‐degree‐of‐freedom (2DOF) Smith predictor based on feedforward compensation method (WO‐2DOFSP‐FC) is proposed. Firstly, the obtained test data of the WRRP are used for system identification and the approximate second‐order model with time‐delay is obtained by using the suboptimal reduction algorithm. This procedure also yields a first‐order model of the disturbance. Then, the whale optimization 2DOF Smith predictor based on feedforward compensation (2DOFSP‐FC) method is proposed. This proposed method improves the 2DOFSP and introduces feedforward compensation control to reduce the influence of the measurable disturbance on the temperature control system. In addition, the Whale Optimization Algorithm (WOA) is proposed to optimally tune the 2DOFSP‐FC controller parameters. Finally, the simulation results show that the proposed WO‐2DOFSP‐FC method can make the controlled system have good characteristics of both set value tracking and disturbance rejection.

Grid-Side Current Harmonic Suppression and Power Factor Improvement Using q – Axis Damping Current Injection for PMSM Drives Without Electrolytic Capacitor

The dc-link film capacitor with low-capacitance in permanent magnet synchronous motor (PMSM) drives without electrolytic capacitor will not only reduce the system stability, but also generate LC resonance with the grid-side filter inductor, which is the important reason for high grid side harmonic current and low power factor. In this paper, a grid side current harmonic suppression strategy using q-axis damping current injection is proposed to address these issues of PMSM drives without electrolytic capacitor. The analysis of the transfer function based on the simplified circuit of the drive shows the instability and LC resonance of the system. The feedforward modification method based on the difference between the quasi-dc-link voltage and the dc-link voltage is applied to improve the system performance. Meanwhile, the specific implementation of the proposed method by injecting damping current into the q-axis is introduced. In addition, a phase disturbance angle is introduced in the proposed control strategy to adjust the phase of quasi-dc-link voltage. By properly adjusting the phase disturbance angle, the conduction angle of the single-phase uncontrolled rectifier bridge can be effectively expanded, and the grid side power factor can be further improved. Experiments on a PMSM drive with the low-capacitance of dc-link film capacitor are carried out to verify the effectiveness of the proposed method.