Model-based Control Research Articles

Highly efficient inverse lumped modeling for the pre-strained circular dielectric elastomer

The equi-biaxially pre-strained circular dielectric elastomer actuator (PCDEA) is a classical actuator configuration to demonstrate the electromechanical deformation mechanism of dielectric elastomer. Till now, there still lacks an effective method to analyze the non-uniform voltage-induced deformation of PCDEA with both high accuracy and low computational cost. Besides, the inverse problem that determines the voltage input for a desired deformation is also a hard problem for the current modeling approaches. In this work, a novel shape function approximation method is proposed to model the non-uniform deformation of PCDEA in a highly accurate lumped-parameter way. The method can solve both the inverse problem and forward problem of the PCDEA, making it possible to position the onset of electromechanical instability (EMI) of this configuration. Detailed comparisons are made between this method and high-resolution finite element analysis to verify its accuracy. This method also features low computational cost, providing opportunities for developing model-based real-time control and proprioception techniques. Further investigation is also carried out on how material types and configurational parameters interact with PCDEAs’ performance, along with checking the findings with experiment observation. Guidelines for DE material development and actuator design are also summarized accordingly.

Dynamic Modeling and Tracking Of 3-Wheels Omni-Directional Mobile Robot with Manual PID Tuning Method

This paper presents the dynamic modelling and tracking control of a holonomic 3-wheel omni-directional mobile robot. A proportional integral derivative (PID) controller is used, and its performance depends on the dynamic model's accuracy. The controller design for the switched non-linear system is based on both a dynamic model and a kinematic design. A proposed dynamic model averages the contact radius of the 3-wheels. The resultant dynamic model, on average, is non-linear and smooth, serving as a potential solution for model-based control design. The closed-loop simulation results clearly illustrate the dynamic characteristics of the 3-wheels omni-directional mobile robot, offering valuable insights into its behaviour for further analysis and optimisation. Comparative simulation results using MATLAB are presented and demonstrate the effectiveness and legitimacy of the control mechanism. Keywords: Holonomic, kinematic, dynamic model, PID controller, non-linear, omni-directional robot

Reinforcement learning based temperature control of a fermentation bioreactor for ethanol production.

Ethanol production is a significant industrial bioprocess for energy. The primary objective of this study is to control the process reactor temperature to get the desired product, that is, ethanol. Advanced model-based control systems face challenges due to model-process mismatch, but Reinforcement Learning (RL) is a class of machine learning which can help by allowing agents to learn policies directly from the environment. Hence a RL algorithm called twin delayed deep deterministic policy gradient (TD3) is employed. The control of reactor temperature is categorized into two categories namely unconstrained and constrained control approaches. The TD3 with various reward functions are tested on a nonlinear bioreactor model. The results are compared with existing popular RL algorithm, namely, deep deterministic policy gradient (DDPG) algorithm with a performance measure such as mean squared error (MSE). In the unconstrained control of the bioreactor, the TD3 based controller designed with the integral absolute error (IAE) reward yields a lower MSE of 0.22, whereas the DDPG produces an MSE of 0.29. Similarly, in the case of constrained controller, TD3 based controller designed with the IAE reward yields a lower MSE of 0.38, whereas DDPG produces an MSE of 0.48. In addition, the TD3 trained agent successfully rejects the disturbances, namely, input flow rate and inlet temperature in addition to a setpoint change with better performance metrics.

Hybrid Advanced Control Strategy for Post-Combustion Carbon Capture Plant by Integrating PI and Model-Based Approaches

Even though the energy penalties and solvent regeneration costs associated with amine-based absorption/stripping systems are important challenges, this technology remains highly recommended for post-combustion decarbonization systems given its proven capture efficacy and technical maturity. This study introduces a novel centralized and decentralized hybrid control strategy for the post-combustion carbon capture plant, aimed at mitigating main disturbances and sustaining high system performance. The strategy is rooted in a comprehensive mathematical model encompassing absorption and desorption columns, heat exchangers and a buffer tank, ensuring smooth operation and energy efficiency. The buffer tank is equipped with three control loops to finely regulate absorber inlet solvent solution parameters, preventing disturbance recirculation from the desorber. Additionally, a model-based controller, utilizing the model predictive control (MPC) algorithm, maintains a carbon capture yield of 90% and stabilizes the reboiler liquid temperature at 394.5 K by manipulating the influent flue gas to the lean solvent flowrates ratio and the heat duty of the reboiler. The hybrid MPC approach reveals efficiency in simultaneously managing targeted variables and handling complex input–output interactions. It consistently maintains the controlled variables at desired setpoints despite CO2 flue gas flow disturbances, achieving reduced settling time and low overshoot results. The hybrid control strategy, benefitting from the constraint handling ability of MPC, succeeds in keeping the carbon capture yield above the preset minimum value of 86% at all times, while the energy performance index remains below the favorable value of 3.1 MJ/kgCO2.

An Improved Onboard Adaptive Aero-Engine Model Based on an Enhanced Neural Network and Linear Parameter Variance for Parameter Prediction

Achieving measurable and unmeasurable parameter prediction is the key process in model-based control, for which an accurate onboard model is the most important part. However, neither nonlinear models like component level models or LPV models, nor linear models like state–space models can fully meet the requirements. Hence, an original ENN-LPV linearization strategy is proposed to achieve the online modelling of the state–space model. A special network structure that has the same format as the state–space model’s calculation was applied to establish the state–space model. Importantly, the network’s modelling ability was improved through applying multiple activation functions in the single hidden layer and an experience pool that records data of past sampling instants, which strengthens the ability to capture the engine’s strongly nonlinear dynamics. Furthermore, an adaptive model, consisting of a component-level model with adaptive factors, a linear Kalman filter, a predictive model, an experience pool, and two ENN-LPV networks, was developed using the proposed linearization strategy as the core process to continuously update the Kalman filter and the predictive model. Simulations showed that the state space model built using the ENN-LPV linearization strategy had a better model identification ability in comparison with the model built using the OSELM-LPV linearization strategy, and the maximum output error between the ENN-LPV model and the simulated engine was 0.1774%. In addition, based on the ENN-LPV linearization strategy, the adaptive model was able to make accurate predictions of unmeasurable performance parameters such as thrust and high-pressure turbine inlet temperature, with a maximum prediction error within 0.5%. Thus, the effectiveness and the advantages of the proposed method are demonstrated.

Deep Reinforcement Learning for sim-to-real policy transfer of VTOL-UAVs offshore docking operations

This paper proposes a novel Reinforcement Learning (RL) approach for sim-to-real policy transfer of Vertical Take-Off and Landing Unmanned Aerial Vehicle (VTOL-UAV). The proposed approach is designed for VTOL-UAV landing on offshore docking stations in maritime operations. VTOL-UAVs in maritime operations encounter limitations in their operational range, primarily stemming from constraints imposed by their battery capacity. The concept of autonomous landing on a charging platform presents an intriguing prospect for mitigating these limitations by facilitating battery charging and data transfer. However, current Deep Reinforcement Learning (DRL) methods exhibit drawbacks, including lengthy training times, and modest success rates. In this paper, we tackle these concerns comprehensively by decomposing the landing procedure into a sequence of more manageable but analogous tasks in terms of an approach phase and a landing phase. The proposed architecture utilizes a model-based control scheme for the approach phase, where the VTOL-UAV is approaching the offshore docking station. In the Landing phase, DRL agents were trained offline to learn the optimal policy to dock on the offshore station. The Joint North Sea Wave Project (JONSWAP) spectrum model has been employed to create a wave model for each episode, enhancing policy generalization for sim2real transfer. A set of DRL algorithms have been tested through numerical simulations including value-based agents and policy-based agents such as Deep Q Networks (DQN) and Proximal Policy Optimization (PPO) respectively. The numerical experiments show that the PPO agent can learn complicated and efficient policies to land in uncertain environments, which in turn enhances the likelihood of successful sim-to-real transfer.

Motion Control of a Hybrid Quadruped-Quadrotor Robot

Multimodal motion capability is an emerging topic in the robotics field, and this paper presents a hybrid robot system maneuvering in both terrestrial and aerial environments. Firstly, a micro quadruped–quadrotor robot with onboard sensing and computing is developed. This robot incorporates both the high mobility of unmanned aerial vehicles and the long endurance of mobile robots on the ground. A coordinated motion control scheme is then exploited for adaptive terrestrial–aerial motion transition. In this scheme, a bio-inspired terrestrial locomotion controller is proposed to generate various quadruped locomotions, and a model-based aerial locomotion controller is proposed to generate various quadrotor configurations. Then, an unified motion controller for the two subsystems which dynamically adjusts crawling and flying motion in a complicated environment is presented. Consequently, several practical trials are conducted to demonstrate the adaptability and the robustness of the proposed system.

Designing, modeling and controlling the angular bending of a foldable soft actuator without using a curvature sensor

Soft actuators have recently attracted considerable attention owing to their inherent flexibility and adaptability. Nevertheless, for a soft robot to successfully engage with its surroundings and perform tasks with optimal effectiveness, it encounters a range of obstacles, including the need for precise and skillful movement, the capacity to perceive its own position and motion and the ability to effectively regulate its flexible structures. Researchers have developed techniques to integrate curvature sensors onto flexible devices, enabling them to detect and react to their positions. However, the integration of curvature sensors into flexible structures presents a substantial challenge in the structural manufacturing process. To address these concerns, this article presents a technique for designing, dynamic modeling and controlling the bending angle of foldable soft actuator without the need for curvature sensors. An optimal design for the geometric dimensions of the soft structure utilizing origami concepts to guarantee the requisite bending properties is suggested. A model-based control method that considers both the motion dynamic and the air dynamic is proposed for controlling the angular bending of the actuator. The motion dynamic was developed using the constant volume principle of the elastomer material and the neo-Hookean hyperelastic theory to establish the correlation between the applied pressure and bending angle. This dynamic model incorporates both the hyperelastic material characteristics of silicone rubber and the geometry of the actuator. Soft actuators have variations in the air chamber’s volume during operation and accurately measuring this variation is challenging. In order to tackle this problem, the fuzzy active disturbance rejection controller is used to predict these variations. The controller possesses exceptional position-tracking capability. This control strategy exhibited excellent responsiveness throughout the range of steady-state error values from approximately 1°–2°. Removing the curvature sensor increases the longevity of this soft actuator and promotes the efficiency of the manufacturing process, hence enhancing the practical application possibilities for the soft actuator made from super elastic material.

Prescribed performance based adaptive model-free control for highly flexible spacecraft detumbling rotating satellites

The method, using a servicing spacecraft (SS) mounted with very flexible operation devices (FOD), has shown promising application in detumbling rotating satellites due to its unique advantage of compliant contact processes. The SS is highly flexible, and successful implementation of detumbling rotating satellites is dependent on the advanced controller with high performance. However, the controller design suffers from two problems: (i) since the uncertain dynamics arising from the FOD lead to difficulty in acquiring an accurate dynamic model of the SS, the conventional model-based controllers are unable to achieve high performance; (ii) the large deformation of the FOD caused by contact processes generates impactive disturbance that severely deteriorates the transient and steady-state performances. To address these problems, this paper proposes a prescribed performance based adaptive model-free control for the first time. In the controller design process, a finite-time prescribed performance function is designed to incorporate into the model-free controller, which ensures the user-specified error constraints concerning the convergence rate, overshoot, and tracking accuracy. Meanwhile, an adaptive estimation algorithm is developed to significantly enhance the robustness against the disturbance. Moreover, an anti-windup compensator is constructed to effectively handle the adverse effect of thruster saturation. Simulation results validate the effectiveness of the proposed controller, and further show that the SS can successfully detumble the rotating satellite while achieving prescribed performance.

A comprehensive dynamic model and configuration control of a free-floating soft manipulator-spacecraft system

Space robots are inseparable components of many space missions. However, capabilities of space robots with rigid manipulators are restricted in confronting with unknown and confined environments and situations require high safety. To overcome these defects, the potential of bio-inspired soft robots for space applications has recently been addressed in the literature. Dynamic modeling and control of soft manipulators are challenging due to their infinite degrees of freedom. In this paper, a general comprehensive three-dimensional dynamic modeling framework is developed for a floating soft manipulator-spacecraft system employing the Euler-Lagrange method. This framework derives the coupled equations of motion adapting the generalized Jacobian approach. Consequently, the opportunity of exploiting model-based control methods will be provided. In addition, the proposed model avoids numerical singularities once the bending angles are near zero. Furthermore, the generalized Jacobian matrix is defined for any arbitrary point of the soft manipulator essential for the contact dynamics. Eventually, to verify the proposed dynamic modeling procedure, the dynamic response of a floating soft manipulator-spacecraft system released from an initial configuration is investigated. Then, a sliding mode controller is designed for the floating soft manipulator to track the desired shape and compared to the proportional-derivative control scheme. The obtained results follow the physical principles and fulfill the control objective.

Effective Nonlinear Predictive and CTC-PID Control of Rigid Manipulators

Effective nonlinear control of manipulators with dynamically coupled arms, like those with direct drives, is the subject of the paper. The main proposal of the paper are model-based predictive control (MPC) algorithms, with nonlinear state-space models and most recent disturbance attenuation technique. This technique makes controller design and calculations simpler, avoiding necessity of dynamic modeling of disturbances or resorting to additional techniques like SMC. The core of the paper are computationally effective MPC-NPL (Nonlinear Prediction and Linearization) algorithms, where computations at every sample are divided into two parts: prediction of initial trajectories using nonlinear model, then optimization using simplified linearized model. For a comparison a known CTC-PID algorithm, which is also model-based, is considered. It is applied in standard form and also proposed in more advanced CTC-PID2dof version. For all algorithms a comprehensive comparative simulation study is performed, for a direct drive manipulator under disturbances. Additional contribution of the paper is an investigation of influence of sampling period length and computational delay time on performance of the algorithms, which is practically important when using model-based algorithms and fast sampling.

An Improved Embedded based PID Controller for Boosting Operation in Solar Energy-Battery Systems

The proposed work introduces an improved PID controller for boosting operation in solar energy-battery systems, leveraging model-based controller (MBC) techniques. In general, boost converters are crucial for managing and optimizing power flow in renewable energy systems. While PID controllers are commonly used in boosting applications due to their simplicity, they often face challenges in adapting to the dynamic and nonlinear characteristics of solar energy systems. This can result in suboptimal performance and efficiency. Hence, a model-based controller with manual tuning is proposed to improve the stability and efficiency by systematically tuning controller gain parameters based on its system model. Extensive simulation studies conducted using MATLAB/Simulink shows that the proposed controller outperforms the conventional PID controller under various operating conditions, such as fluctuations in solar irradiance and changes in battery state of charge. Further, the experimental tests conducted on a prototype setup validates the feasibility and ability of the model-based controller to accurately regulate the solar panel output and efficiently charge the battery, which leads to enhanced performance. This work offers a promising solution for enhancing energy conversion efficiency and facilitates the effective integration of renewable energy sources into the power grid.

Neural network PID control of spherical permanent magnet based on external magnetic field driving system

Abstract The neural network PID control method of the external magnetic field driving system solves the shortcomings of the traditional control method. The controlled object is a spherical permanent magnet. Other detections or operation equipment can be attached to the sphere. In the process of external magnetic field driving, the model-based controller is prone to large fluctuations when the environmental parameters change, which reduces the control performance. The neural network algorithm has good control ability for the highly nonlinear magnetic levitation driving system. PID control system combined with a neural network algorithm has high stability and strong adaptability in the driving process. The test results show that the response output overshoot is 9.322%, and the system fluctuation is low. Self-tuning and adaptive environment changes of PID controller parameters can be realized by neural network control.

Nonlinear model-based predictive control of a yeast fermentation bioreactor using new bee colony-WNN modelling structure

This paper concentrates on the application of a newly developed model for the nonlinear control of a yeast fermentation bioreactor. The primary control objectives are the maintenance of temperature and product concentration within the desired range. So, having an accurate model in control strategy is not ignorable and this model influences on close-loop controllability. In the context of online control, the parameters are changed (the system is time varying) and precision of extracted model is more important. Moreover, there is a demand for a computationally efficient model, and the convergence speed during the training of the neural network is significant. To overcome the challenges associated with slow convergence, the risk of being trapped in local minima, and the imperative of achieving computational efficiency, this study introduces a novel combination of a pruned Bee Colony Wavelet Neural Network for the online modeling of the bioreactor. The method's superiority lies in its ability to conduct searches in multiple sections independently, thereby avoiding local minima. The hybrid model controllability is further investigated using Dynamic Matrix Control (DMC) and Generalized Matrix Control (GPC) techniques. Subsequently, the paper explores the application of Nonlinear Model Predictive Control (NMPC) by solving a Gauss-Newton quadratic programming problem online, followed by a comparative analysis of the results. It is demonstrated that the precision of the model closely influences profitability and enhances product yields.

Physical Model-Based VDCM Control to Enhance the Inertia of DC Bus for PV-BESS Supported Off-Board Charging Station

Physical Model-Based VDCM Control to Enhance the Inertia of DC Bus for PV-BESS Supported Off-Board Charging Station

Improvements to a novel learning algorithm for model-based wind turbine controllers

Wind turbine controllers are nowadays ever more advanced and rely on accurate internal controller model information. Therefore a calibrated model is needed for attaining predictable controller performance and ensuring stable operation. To calibrate the internal model information, a novel learning control scheme has recently been proposed that exploits the dynamics of the closed-loop controlled wind turbine system, without the need for wind speed measurements. The learning algorithm thereby periodically excites the generator power controller input signal. An extremum-seeking demodulation scheme was used to calibrate the internal model information. This paper improves the existing learning scheme in two ways: Firstly, it investigates how the frequency of the excitation signal influences the signal-to-noise ratio. Secondly, the problem was reformulated as a root-finding problem. This requires using the in-phase component of the phase-corrected learning signal. In addition, a precalculated lookup table relates the measured in-phase component directly to model uncertainty. It was found that an increased excitation frequency improves the signal-to-noise ratio (SNR) by an order of magnitude. Combined, these contributions improve the convergence speed more than twenty times, addressing the effect of aerodynamic degradation and its consequences on controller performance.

A complete dynamic modeling of two-by-one gas-steam combined cycle unit for coordinated control design

The high penetration of renewable energy can potentially change the role of combined cycle gas turbines (CCGT) from load suppliers to flexible suppliers. However, the slow process in the dynamic modeling of the CCGT complete system restricts the design of its best flexible operation strategy. Therefore, this paper adopts the composite modeling method to establish a complete nonlinear dynamic model of a 2 × 1 CCGT including two gas turbines, two three-pressure heat recovery steam generators (HRSG), one steam turbine, and one heat exchanger. This model can reflect the dynamic characteristics of pressure and flow in the bottom cycle due to the changes in operating mode, which can guide the design of the flexible operation control strategy of the 2 × 1 CCGT. To verify the model’s accuracy, the simulated data were compared with the field data from a 900 MW cycle with the same configuration. The results show that the established model can accurately reflect the changing trend of variables during mode switching. The range errors of essential variables such as gas turbine (GT) power is 3.8 %, GT outlet temperature is 1.4 %, high-pressure drum pressure is 4.5 %, and the range errors of other variables are below 5 %, proving the fidelity of the established control model. The dynamic model established in this paper will be the basis for model-based controller design which can handle multi-objective optimization and constraints in flexible operations. Moreover, this model will be used to guide the design of multi-mode control and energy management strategies.

Combining learning and control in linear systems

In this paper, we provide a theoretical framework that separates the control and learning tasks in a linear system. This separation allows us to combine offline model-based control with online learning approaches and thus circumvent current challenges in deriving optimal control strategies in applications where a large volume of data is added to the system gradually in real time and not altogether in advance. We provide an analytical example to illustrate the framework.

Development of K-Maryblyt for Fire Blight Control in Apple and Pear Trees in Korea.

K-Maryblyt has been developed for the effective control of secondary fire blight infections on blossoms and the elimination of primary inoculum sources from cankers and newly emerged shoots early in the season for both apple and pear trees. This model facilitates the precise determination of the blossom infection timing and identification of primary inoculum sources, akin to Maryblyt, predicting flower infections and the appearance of symptoms on various plant parts, including cankers, blossoms, and shoots. Nevertheless, K-Maryblyt has undergone significant improvements: Integration of Phenology Models for both apple and pear trees, Adoption of observed or predicted hourly temperatures for Epiphytic Infection Potential (EIP) calculation, incorporation of adjusted equations resulting in reduced mean error with 10.08 degree-hours (DH) for apple and 9.28 DH for pear, introduction of a relative humidity variable for pear EIP calculation, and adaptation of modified degree-day calculation methods for expected symptoms. Since the transition to a model-based control policy in 2022, the system has disseminated 158,440 messages related to blossom control and symptom prediction to farmers and professional managers in its inaugural year. Furthermore, the system has been refined to include control messages that account for the mechanism of action of pesticides distributed to farmers in specific counties, considering flower opening conditions and weather suitability for spraying. Operating as a pivotal module within the Fire Blight Forecasting Information System (FBcastS), K-Maryblyt plays a crucial role in providing essential fire blight information to farmers, professional managers, and policymakers.

Robust Geometric Control for a Quadrotor UAV with Extended Kalman Filter Estimation

This study proposes a robust geometric controller tailored for quadrotor unmanned aerial vehicles (UAVs). The original geometric controller exhibits excellent performance in quadrotor UAV maneuvers. However, as a model-based nonlinear control method, it is sensitive to system model parameters. By integrating a novel extended Kalman filter (EKF)-based estimator for real-time, online estimation of the quadrotor’s inertia parameters, the controller adeptly handles internal uncertainties and external perturbations during flight maneuvers. This approach significantly improves the robustness of the control system against model inaccuracies. Empirical evidence is provided through both simulation and extensive real-world flight tests, demonstrating the controller’s effectiveness and its practical applicability in dynamic environments. The results confirm that this integration substantially enhances system reliability and performance under varied operational conditions.