Online Parameter Research Articles

Deep Q-Network-Enhanced Self-Tuning Control of Particle Swarm Optimization

Particle Swarm Optimization (PSO) is a widespread evolutionary technique that has successfully solved diverse optimization problems across various application fields. However, when dealing with more complex optimization problems, PSO can suffer from premature convergence and may become stuck in local optima. The primary goal is accelerating convergence and preventing solutions from falling into these local optima. This paper introduces a new approach to address these shortcomings and improve overall performance: utilizing a reinforcement deep learning method to carry out online adjustments of parameters in a homogeneous Particle Swarm Optimization, where all particles exhibit identical search behaviors inspired by models of social influence among uniform individuals. The present method utilizes an online parameter control to analyze and adjust each primary PSO parameter, particularly the acceleration factors and the inertia weight. Initially, a partially observed Markov decision process model at the PSO level is used to model the online parameter adaptation. Subsequently, a Hidden Markov Model classification, combined with a Deep Q-Network, is implemented to create a novel Particle Swarm Optimization named DPQ-PSO, and its parameters are adjusted according to deep reinforcement learning. Experiments on different benchmark unimodal and multimodal functions demonstrate superior results over most state-of-the-art methods regarding solution accuracy and convergence speed.

Closed-loop control strategy design for Caputo–Fabrizio definition-based fractional-order Boost converters considering parametric uncertainties

This paper proposes a dual closed-loop adaptive PI control system based on parameter identification for Caputo–Fabrizio (C-F) definition-based fractional-order boost converters. First, a C-F definition-based fractional-order model of the Boost converter is derived. Afterward, an online component parameter identification method is proposed to obtain the values and orders of capacitors and inductors in the proposed model. Based on these identified parameters, a dual closed-loop adaptive PI control system considering parametric uncertainties for the Boost converters is designed using the frequency domain analysis method. Finally, simulation validations are conducted. The online parameter identification method can accurately identify the values and orders of capacitors and inductors. The identification error is less than [Formula: see text] for the capacitance value and [Formula: see text] for the other parameters. Moreover, the proposed control system can automatically adjust the parameters with changes in the component parameters, effectively improving the robustness. Compared to traditional PI controllers, when the parameters of components change, the proposed control system can reduce voltage overshoot by [Formula: see text] in the case of sudden changes in the input voltage. These results confirm the efficacy of the proposed online parameter identification method and the dual closed-loop adaptive PI control system.

Real-time correction of gain nonlinearity in electrostatic actuation for whole-angle micro-shell resonator gyroscope

MEMS gyroscopes are well known for their outstanding advantages in Cost Size Weight and Power (CSWaP), which have inspired great research attention in recent years. A higher signal-to-noise ratio (SNR) for MEMS gyroscopes operating at larger vibrating amplitudes provides improved measuring resolution and ARW performance. However, the increment of amplitude causes strong nonlinear effects of MEMS gyroscopes due to their micron size, which has negative influences on the performance. This paper carries out detailed research on a general nonlinear mechanism on the sensors using parallel-plate capacitive transducers, which is called the gain nonlinearity in electrostatic actuation. The theoretical model established in this paper demonstrates the actuation gain nonlinearity causes the control-force coupling and brings extra angle-dependent bias with the 4th component for the whole-angle gyroscopes, which are verified by the experiments carried out on a micro-shell resonator gyroscope (MSRG). Furthermore, a real-time correction method is proposed to restore a linear response of the electrostatic actuation, which is realized by the gain modification with an online parameter estimation based on the harmonic-component relationship of capacitive detection. This real-time correction method could reduce the 4th component of the angle-dependent bias by over 95% from 0.003°/s to less than 0.0001°/s even under different temperatures. After the correction of actuation gain nonlinearity, the bias instability (BI) of whole-angle MSRG is improved by about 3.5 times from 0.101°/h to 0.029°/h and the scale factor nonlinearity (SFN) is reduced by almost one order of magnitude from 2.02 ppm to 0.21 ppm.

Adaptive robust motion control for hydraulic load sensitive systems considering displacement dynamic compensation

Hydraulic load-sensitive systems (HLSS) are widely used for high power density and energy efficiency. This study introduces an adaptive, energy-efficient HLSS with a valve-controlled variable motor. The system faces challenges from non-linearities, including internal higher-order dynamics due to displacement changes and external unknown disturbances, which hinder precision applications. To address this issue, this study explores HLSS principles to develop an accurate system model. Subsequently, an adaptive robust motion control that considers displacement compensation (DCARC) is proposed using the established model. DCARC can learn unknown parameters online and compensate the model more accurately to improve control accuracy. Experiments show that considering the higher order dynamic effects caused by displacement in the system can improve model accuracy and effectively reduce the burden of parameter adaptation and robust feedback terms. High-precision and energy-efficient HLSS motion is verified and realized in the study. The control accuracy of DCARC is 19.4% higher than that of conventional adaptive robust control (ARC). Under experimental conditions, the proposed system can improve energy efficiency by up to five times compared to valve-controlled fixed displacement motor systems (VFDS).

Adaptive fuzzy neural super-twisting control of micro gyroscope sensor

To maintain the vibrations of the gyroscope proof mass, an adaptive super-twisting sliding mode control (STSMC) of a micro gyroscope based on a two-loop recursive fuzzy neural network (TLRFNN) is designed. In order to estimate the unknown parameters of the nonlinear system online, an adaptive technology is adopted based on the parameter linearization of the nonlinear model. In response to the problem of system chattering caused by ordinary sliding mode control, the STSMC with the advantages of high order sliding mode and conventional sliding mode is introduced. Aiming at the problem of uncertainty in system parameters, a two-loop recursive fuzzy neural network approximation with the competence over the storage of a priori information is utilized. All adaptive laws are obtained under the Lyapunov stability framework which contributes to the stability of the system. The simulation research shows the system exhibits good tracking performance and maintains a small tracking error under the control of the suggested control system. The effect is measured by calculating the root mean square error (RMSE) parameter of the tracking error and the proposed control system achieves a tracking effect of \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${\ ext{1}}{\ ext{0.8590}} \ imes {\ ext{1}}{{\ ext{0}}^{ - 4}}$$\\end{document} and \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${\ ext{2}}{\ ext{0.0831}} \ imes {\ ext{1}}{{\ ext{0}}^{ - 4}}$$\\end{document} in the x and y directions, respectively.

State and parameter estimation of a dynamic froth flotation model using industrial data

This paper investigates an observable dynamic model of froth flotation circuits aimed at online state and parameter estimation and model-based control. The aim is to estimate the model states and parameters online from industrial data. However, in light of limitations in the plant data, additional model analysis is conducted. It is shown that without online compositional measurements, only the states and parameters of a reduced model can be estimated online. The reduced model lumps all recovery mechanisms into a single empirical equation. The reduced model is used to develop a moving horizon estimator (MHE) which is implemented on the industrial data. The state and parameter estimates from the MHE are used to evaluate the model prediction accuracy over a receding control horizon as would be done in model predictive control (MPC). Given the uncertainty of the available data, unmeasured disturbances and missing online measurements, the estimation and prediction results are reasonably accurate, at least in a qualitative sense. If accurate and reliable online measurements are available for estimation, the reduced model shows potential to be used for long-term model-based supervisory control of a flotation circuit.

Improved online parameter identification coupled with discrete wavelet transform for state-of-charge estimation under noise-riding discharging/charging battery voltage

With the expansion of the battery market and electric vehicles, improved battery-state estimation approaches are required for energy management to meet the requirements for safe usage. The accurate identification of battery parameters significantly influences the state-of-charge (SOC) computation algorithms. In addition, noise from charging and discharging voltages can influence the accuracy of the SOC estimation. This paper proposes a novel methodology combining a discrete wavelet transform (DWT) denoising procedure and an adaptive parameter identification. The DWT uses a multi-resolution analysis (MRA) to decompose noise-riding charging/discharge voltage signals, which are then rebuilt using the inverse DWT (IDWT) based on the rigsure adaptive threshold selection rule. In this study, a dual-resistor-capacitor lithium-ion battery model was examined. A formulation of the modified forgetting factor recursive least squares (MFFRLS) was presented and demonstrated robustness in overcoming data saturation while enhancing parameter tracking ability even under the effect of noise. To verify the effectiveness of the coupled methods for SOC estimation defined in this study, experimental data obtained from various working conditions were employed to estimate the SOC using the extended Kalman filter (EKF). A thorough analysis demonstrated that the proposed MFFRLS combined with the denoising method performed well in terms of SOC estimation accuracy and robustness.

Degradation modeling of serial space lithium-ion battery pack based on online inconsistency representation parameters

Establishing an inconsistency-based degradation model for lithium-ion battery packs is crucial for suppressing the degradation of battery packs by optimizing the inconsistency. This paper proposes a method for modeling the degradation of serial space lithium-ion battery packs based on online inconsistency representation parameters. Firstly, the static inconsistency representation parameters are acquired online by quantifying the difference among voltage intervals of cells through Gaussian distribution, addressing the challenge of acquiring static inconsistency representation parameters online. Simultaneously, dynamic inconsistency representation parameters are serialized representations by quantifying the voltage differences of cells at multiple moments, better capturing the rapid evolutionary process of inconsistency. Secondly, a linear model is used to model the serialized parameters and degradation, simplifying the modeling process while achieving multi-source information fitting. Meanwhile, a nonlinear model is constructed to better fit the nonlinear degradation trend of the battery pack. Then, Kalman filter is utilized for model fusion to achieve complementary advantages and improve modeling accuracy. Cross-validation with laboratory battery pack data confirms that this method outperforms comparison approaches, achieving the mean absolute error and maximum error of less than 1.73 % and 3.31 %, respectively. The method proposed in this paper provides a basis for future work on battery pack life extension.

Vehicle posture wheelbase preview control of in-wheel motors drive intelligent vehicle on potholed road

The in-wheel motors drive intelligent vehicle (IMDIV) has a strong ability of dynamic control, which makes it be a major development path of intelligent vehicles in the future. However, its stability is easily deteriorated by the impact of the road when driving on potholed road. To tackle this issue, a vehicle posture control method based on road feedback and elevation recognition is proposed. Firstly, an elevation recognition algorithm of adaptive Kalman filter (AKF) based on vehicle dynamic response is proposed after considering the flooded road and the system noise uncertainty. Secondly, a vehicle posture controller of fuzzy active disturbance rejection control (FADRC) is designed, with the function of dynamically adjusting the active disturbance rejection control (ADRC) parameters. It is designed to implement front suspension feedback control and rear suspension wheelbase preview control. At last, the road elevation recognition algorithm and vehicle posture control method are proved by Simulink simulation and off-line hardware in the loop test. The results revealed that the AKF can effectively recognize both white noise pavement and potholed road, and the recognition accuracy is greater than 90%. The FADRC controller realizes accurate front suspension feedback control and rear suspension wheelbase preview control by adjusting the real-time online parameters of ADRC. The proposed vehicle posture control method can effectively control the pitch and roll motion, which significantly improves the stability of IMDIV on potholed road. This study can serve as a reference for the posture stability control of IMDIV on potholed road.

Sensorless Human–Robot Interaction: Real‐Time Estimation of Co‐Grasped Object Mass and Human Wrench for Compliant Interaction

Human–robot physical interaction in shared object manipulation can fully leverage the strengths of both humans and collaborative robots, achieving better interaction outcomes. However, in typical physical interactions, the inertia parameters of the object are either assumed to be known or ignored, simplifying the interaction to be directly between the human and the collaborative robot. This simplification can lead to inaccuracies in the robot's understanding of human intent due to deviations in object dynamics. This article presents a method that eliminates the need for a force sensor on the human side during human–robot collaboration. Using an extended Kalman filter, the method estimates object parameters online and applies a disturbance observer to determine human force. This approach reduces human effort and improves tracking performance compared to traditional methods. It also avoids the hassle of installing and removing sensors in complex scenarios. By estimating object mass and human force in real time, the robot can adjust its behavior to ensure safer and smoother interactions. Furthermore, the system can handle various objects without requiring specific programming for each case.

Obstacle avoidance shape control of deformable linear objects with online parameters adaptation based on differentiable simulation

The manipulation of deformable linear objects (DLOs) such as ropes, cables, and hoses by robots has promising applications in various fields such as product assembly and surgical suturing. However, DLOs are more difficult to manipulate than rigid objects because their shape changes during manipulation. Furthermore, preventing a DLO from colliding with the environment is important to prevent it from becoming entangled and causing shape control to fail. In this paper, we proposed an obstacle avoidance and shape control scheme for DLOs based on differentiable simulation that does not require prior data or a specialized controller. First, we established a dynamic model of the DLO that allows for both forward dynamics transfer and error backpropagation to obtain gradients. Then, we employed model predictive control to optimize the embedded neural network for predicting the actions that would manipulate the DLO. Finally, the control scheme was made applicable to DLOs with different material properties by allowing online adaptation of the model parameters essential to deformation during manipulation. Simulations and real-world experiments demonstrate that the proposed control scheme could manipulate the DLO stably and accurately to avoid obstacles and achieve the goal state. In addition, the online adaptation of parameters helped mitigate the sim-to-real gap.

Identifying the parameters of ultracapacitors based on variable forgetting factor recursive least square

PurposeUltracapacitors find extensive applications in various fields because of their high energy density and long cycling periods. However, due to the movement of ions and the arrangement patterns on rough/irregular electrode surfaces during the charge and discharge process of ultracapacitors, the parameters of ultracapacitors usually change with the variation of operating conditions. The purpose of this study is to accurately and quickly identify the parameters of ultracapacitors.Design/methodology/approachA variable forgetting factor recursive least square (VFFRLS) algorithm is proposed in this paper for online identifying the equivalent series resistance and capacitance C of ultracapacitors. In this work, a real-time error-based strategy is developed to adaptively regulate the value of the forgetting factor of traditional forgetting factor recursive least square (FFRLS) algorithm. The strategy uses the square of the average time autocorrelation estimation of the prior error and the posterior error between the predicted output and the actual output as the adjustment basis of forgetting factors.FindingsExperiments were conducted using the proposed scheme, and the results were compared with the estimation results obtained by the recursive least squares (RLS) algorithm and the traditional FFRLS algorithm. The maximum root mean square error between the estimated values and actual values for VFFRLS is 3.63%, whereas for FFRLS it is 9.61%, and for RLS it is 19.33%.Originality/valueBy using the proposed VFFRLS algorithm, a relatively high precision can be achieved for the online parameter estimation of ultracapacitors. Besides, the dynamic balance between parameter stability and tracking performance can be validated by dynamically adjusting the forgetting factor.

Online control parameter optimization design for multi-machine coordinated loading system of hazardous substances

To address the parameter instability issues in hazardous materials handling during multi-machine loading and unloading operations, we propose a Full-Scale Smart Parameter Optimization Control (FSPOC) system specifically designed for multi-machine coordination. This system leverages a novel fish scale prediction algorithm tailored for cooperative multi-machine environments. Initially, the fish scale prediction algorithm, inspired by bionic fish scales, is developed to predict future system behavior by analyzing historical data. Building on this algorithm, we introduce a disturbance cancellation control theorem and design a parameter optimization controller to enhance stability in high-dimensional nonlinear spaces. The FSPOC method is then applied to a multi-machine cooperative system, enabling online distributed parameter optimization for complex systems with multiple degrees of freedom. The effectiveness of the proposed method was validated through simulations, where it was compared with two other optimization techniques: Genetic Algorithm-based PID (GAPID) and Chaotic Atomic Search Algorithm-based PID (CHASO). The simulation results confirm the superiority of the FSPOC method.

Intelligent Control System for Brain-Controlled Mobile Robot Using Self-Learning Neuro-Fuzzy Approach.

Brain-computer interface (BCI) provides direct communication and control between the human brain and physical devices. It is achieved by converting EEG signals into control commands. Such interfaces have significantly improved the lives of disabled individuals suffering from neurological disorders-such as stroke, amyotrophic lateral sclerosis (ALS), and spinal cord injury-by extending their movement range and thereby promoting self-independence. Brain-controlled mobile robots, however, often face challenges in safety and control performance due to the inherent limitations of BCIs. This paper proposes a shared control scheme for brain-controlled mobile robots by utilizing fuzzy logic to enhance safety, control performance, and robustness. The proposed scheme is developed by combining a self-learning neuro-fuzzy (SLNF) controller with an obstacle avoidance controller (OAC). The SLNF controller robustly tracks the user's intentions, and the OAC ensures the safety of the mobile robot following the BCI commands. Furthermore, SLNF is a model-free controller that can learn as well as update its parameters online, diminishing the effect of disturbances. The experimental results prove the efficacy and robustness of the proposed SLNF controller including a higher task completion rate of 94.29% (compared to 79.29%, and 92.86% for Direct BCI and Fuzzy-PID, respectively), a shorter average task completion time of 85.31 s (compared to 92.01 s and 86.16 s for Direct BCI and Fuzzy-PID, respectively), and reduced settling time and overshoot.

Intelligent frequency control of AC microgrids with communication delay: An online tuning method subject to stabilizing parameters

Smart control techniques have been implemented to address fluctuating power levels within isolated microgrids, mitigating the risk of unstable frequencies and the potential degradation of power supply quality. However, a challenge lies in the fact that employing these computationally complex methods without stability preservation might not suffice to handle the rapid changes of this highly dynamic environment in real-world scenarios over communication delays. This study introduces a flexible real-time approach for the frequency control problem using an artificial neural network (ANN) constrained to stabilized regions. Our solution integrates stabilizing PID controllers, computed through small-signal analysis and tuned via an automated search for optimal ANN weights and reinforcement learning (RL)-based selected constraints. First, we design stabilizing PID controllers by applying the stability boundary locus method and the Mikhailov criterion, specifically addressing communication delays. Next, we refine the controller parameters online through an automated process that identifies optimal coefficient combinations, leveraging a constrained ANN to manage frequency deviations within a restricted parameter range. Our approach is further enhanced by employing the RL technique, which trains the tuning system using an interpolated stability boundary curve to ensure both stability and performance. This one-of-a-kind combination of ANN, RL, and advanced PID tuning methods is a big step forward in how we handle frequency control problems in isolated AC microgrids. The experiments show that our solution outperforms traditional methods due to its reduced parameter search space. In particular, the proposed method reduces transient and steady-state frequency deviations more than semi- and unconstrained methods. The improved metrics and stability analysis show that the method improves system performance and stability under changing conditions.

Enhanced Second-Order RC Equivalent Circuit Model with Hybrid Offline–Online Parameter Identification for Accurate SoC Estimation in Electric Vehicles under Varying Temperature Conditions

Accurate estimation of State-of-Charge (SoC) is essential for ensuring the safe and efficient operation of electric vehicles (EVs). Currently, second-order RC equivalent circuit models do not account for the influence of battery charging and discharging states on battery parameters. Additionally, offline parameter identification becomes inaccurate as the battery ages. Online identification requires real-time parameter updates during the SoC estimation process, which increases the computational complexity and reduces the computational efficiency of real vehicle Battery Management System (BMS) chips. To address these issues, this paper proposes a SoC estimation method that combines online and offline identification based on an optimized second-order RC equivalent circuit model, which distinguishes it from existing methods in the field. On the basis of the traditional second-order RC model, the Ohmic resistance (R0), polarization resistance (R1), polarization capacitance (C1), diffusion resistance (R2), and diffusion capacitance (C2) during the charging and discharging processes are discussed separately. R0, which does not change frequently, is identified offline, while R1, R2, C1, and C2, which dynamically change with time and current, are identified online. To thoroughly verify the feasibility of the proposed method, we construct an SoC estimation test bench, which allows us to adjust the battery’s surface temperature in real time using a temperature control chamber. Experimental validation under Federal Urban Driving Schedule (FUDS) (−10 °C to 45 °C, 80% battery capacity) and Dynamic Stress Test (DST) (−10 °C to 45 °C, 8% battery capacity) conditions demonstrate that our method improves SoC estimation accuracy by 16.28% under FUDS and 28.2% under DST compared to the improved GRU-based transfer learning method, while maintaining system SoC estimation efficiency.

State of Charge Estimation of Lithium-Ion Batteries Based on Fractional-Order Model with Mul-ti-Innovations Dual Cubature Kalman Filter Method

An accurate estimation of the lithium battery’s state of charge (SOC) is critical. The article proposes a dual fractional order multi-innovations cubature Kalman filter (DFOMICKF) algorithm for estimating lithium battery SOC. The algorithm adopts the idea of multiple time scales, where one of the FOMICKF is used to identify the circuit model parameters online in the macro time scale. Another FOMICKF is used to estimate the SOC in the micro time scale, and the circuit parameters updated online in real-time are passed into the estimation of the SOC filter to form an online joint estimation method of SOC and circuit parameters. Finally, multiple algorithms of DFOMICKF, FOMICKF, FOCKF, and CKF are compared and experimented under different working conditions to compare and analyze the estimated SOC errors. It is verified that the proposed algorithm can solve the problems of inaccuracy, poor convergence, and poor robustness of the traditional Kalman filtering algorithm for estimating SOC, which has good research value.

Hierarchical Weight Averaging for Deep Neural Networks.

Despite simplicity, stochastic gradient descent (SGD)-like algorithms are successful in training deep neural networks (DNNs). Among various attempts to improve SGD, weight averaging (WA), which averages the weights of multiple models, has recently received much attention in the literature. Broadly, WA falls into two categories: 1) online WA, which averages the weights of multiple models trained in parallel, is designed for reducing the gradient communication overhead of parallel mini-batch SGD and 2) offline WA, which averages the weights of one model at different checkpoints, is typically used to improve the generalization ability of DNNs. Though online and offline WA are similar in form, they are seldom associated with each other. Besides, these methods typically perform either offline parameter averaging or online parameter averaging, but not both. In this work, we first attempt to incorporate online and offline WA into a general training framework termed hierarchical WA (HWA). By leveraging both the online and offline averaging manners, HWA is able to achieve both faster convergence speed and superior generalization performance without any fancy learning rate adjustment. Besides, we also analyze the issues faced by the existing WA methods, and how our HWA addresses them, empirically. Finally, extensive experiments verify that HWA outperforms the state-of-the-art methods significantly.

State of Charge Estimation in Batteries for Electric Vehicle Based on Levenberg–Marquardt Algorithm and Kalman Filter

A new optimization method for estimating the State of Charge (SOC) of battery charge state is proposed. This method incorporates the Levenberg–Marquardt Algorithm (LMA) for online parameter identification and the Extended Kalman Filter (EKF) for SOC. On the one hand, the LMA efficiently alleviates the ’Data saturation’ problem experienced by least squares methods by dynamically adjusting weights of data. On the other hand, the EKF improves the robustness and adaptability of SOC estimation. Simulation results under Hybrid Pulse Power Characteristic (HPPC) conditions demonstrate that this new approach offers superior performance in SOC estimation in batteries for electric vehicles compared to existing methods, with better tracking of the true SOC curve, reduced estimation error, and improved convergence.

Physics and data‐driven approach for online joint state and parameter estimation of electricity and steam networks

Physics and data‐driven approach for online joint state and parameter estimation of electricity and steam networks