Ship Formation Control Using Nonlinear Model Predictive Control With Safe Speed Constraints and Tidal Elevation Variations

Hydrogen is crucial for achieving SDGs by driving energy transition and combating climate change. Proton exchange membrane fuel cell technology, leveraging hydrogen, faces challenges in meeting high‐power demands. The multistack fuel cell system (MFCS) tackles this by integrating multiple substacks, yet its air supply needs meticulous control. Proportional integral derivative (PID) decoupling from single‐stack falls short of MFCS. This article proposes nonlinear model predictive control (NMPC) for optimized air flow and pressure decoupling. Modeling MFCS's air system and designing a predictive model, it is aimed to ensuring precise control of air flow and pressure in each substack. The decoupling experiments show that NMPC outperforms PID, accurately managing air flow and pressure and reducing load fluctuations. For air mass flow, NMPC cuts mean‐absolute error (MAE) by 64.56% and root‐mean‐square error (RMSE) by 81.36%. For pressure, MAE drops 81.23% and RMSE 83.59%. Comprehensive step load tests confirm NMPC's precise, dynamic regulation too, compared to PID, NMPC lowers average MAE for air mass by 20.67%, pressure by 32.22%. RMSE improvements of 31.08% and 33.23% highlight NMPC's strength. NMPC's quick response mitigates coupling issues, enhancing vehicle load adaptability.

Nonlinear Model Predictive Control approach in design of Adaptive Cruise Control with automated switching to cruise control

For omnidirectional mobile robot (OMR), it is difficult to achieve adequate tracking flexibility and accuracy with nonlinear model predictive control (NMPC) in a fixed horizon due to the dynamic change of system variables. This paper proposed a strategy of NMPC with self-turned prediction horizon, which can adapt to the change of system variables and perform a stable tracking control. Firstly, we construct a kinematics model for the Ackerman mode of the OMR. Secondly, the NMPC strategy is adapted to trajectory tracking. Moreover, a method to select the appropriate prediction of horizon is proposed by considering the influence of velocity and road curvature on the system. And the switch problem of the self-turned prediction horizon is solved through the division of the terminal region. Theoretical analysis reveals that the NMPC method is stable, and the convergence of the state errors can be guaranteed. Finally, simulation experiments on a four-wheeled OMR validate the effectiveness and superiority of the self-turned prediction horizon control method comparing with the fixed horizon NMPC method or other variable prediction horizon NMPC methods.

Designing non-linear model predictive control (NMPC) laws for controlling multiple autonomous surface vessels in arbitrary formations in environments containing obstacles are reported in this paper. Two leader-follower decentralized geometrical control schemes that are required for defining a unique two-dimensional formation are considered. A three-degree-of-freedom dynamic model of surface vessels has been used for the controller design. It is assumed that the surface vessels are under-actuated and obstacles are present in the vessels' work environment. The real-time optimization abilities of the NMPC method has been used to improve the response of the unactuated DOF of the vessels and to directly incorporate the local obstacle avoidance into the formation control eliminating the need for an external local obstacle avoidance algorithm. This leads to more effective obstacle avoidance decisions based on the dynamics of the vehicle. The effectiveness the developed control law, even in the presence of model uncertainty and external disturbances are demonstrated via computer simulations.

This paper presents a nonlinear model predictive control (NMPC) method with adaptive neuro-modelling for redundant robotic manipulators. Using the NMPC, the end-effector of the robot tracks a predefined geometry path in the Cartesian space without colliding with obstacles in the workspace and at the same time avoiding singular configurations of the robot. Furthermore, using the neural network for the model prediction, no knowledge about system parameters is necessary; hence, yielding robustness against changes in parameters of the system. Numerical results for a 4DOF redundant spatial manipulator actuated by DC servomotors shows effectiveness of the proposed method.

Real-time random reference tracking nonlinear model predictive control: a case study on wind turbines.

In this work a nonlinear model predictive control method for a floating wind turbine is presented. A reduced nonlinear model including disturbance preview of wind and waves is derived and implemented to compute optimal input trajectories for collective pitch and the generator torque. A cost functional is introduced which fulfills all desired constraints and controller goals for above rated wind conditions. The controller is tested for extreme and fatigue load cases and a significant reduction of the power and rotor speed deviations is obtained. Furthermore, ultimate tower loads and damage equivalent loads on shaft and blades are decreased. Although more detailed testing is necessary, this preliminary results show the advantages of nonlinear model predictive control for floating wind turbines.

Preconditioning for continuation model predictive control

Mobile medical robots have been widely used in various structured scenarios, such as hospital drug delivery, public area disinfection, and medical examinations. Considering the challenge of environment modeling and controller design, how to achieve the information from the human demonstration in a structured environment directly arouse our interests. Learning skills is a powerful way that can reduce the complexity of algorithm in searching space. This is especially true when naturally acquiring new skills, as mobile medical robot must learn from the interaction with a human being or the environment with limited programming effort. In this article, a learning scheme with nonlinear model predictive control (NMPC) is proposed for mobile robot path tracking. The learning-by-imitation system consists of two levels of hierarchy: in the first level, a multivirtual spring-dampers system is presented for imitation of the mobile robot's trajectories; and in the second level, the NMPC method is used in the motion control system. The NMPC strategy utilizes a varying-parameter one-layer projection neural network to solve an online quadratic programming optimization via iteration over a limited receding horizon. The proposed algorithm is evaluated on a mobile medical robot with an emulated trajectory in simulation and three scenarios used in the experiment.

In this paper, the stabilization of a continuous time-delayed system is considered. To control the bifurcation and chaos in a time-delayed system, a parameter perturbation control and a hybrid control are proposed. Then, to ensure the asymptotic stability of the system in the presence of unexpected system parameter changes, the adaptive control idea is introduced, i.e., the perturbation control parameter and the hybrid control parameter are automatically tuned according to the adaptation laws, respectively. The adaptation algorithms are constructed based on the Lyapunov-Krasovskii stability theorem. The adaptive parameter perturbation control and the adaptive hybrid control methods improve the corresponding constant control methods. They have the advantages of increased stability, adaptability to the changes of the system parameters, control cost saving, and simplicity. Numerical simulations for a well-known chaotic time-delayed system are performed to demonstrate the feasibility and superiority of the proposed control methods. A comparison of the two adaptive control methods is also made in an experimental study.

<div class="section abstract"><div class="htmlview paragraph">Platooning is a promising technology which can mitigate greenhouse gas impacts and reduce transportation energy consumption. Platooning is a coordinated driving strategy where trucks align themselves in order to realize aerodynamic benefits to reduce required motive force. The aerodynamic benefit is seen as either a “pull” effect experienced by the following vehicles or a “push” effect experienced by the leader. The energy savings magnitude increases nonlinearly as headway (following distance) is reduced [<span class="xref">1</span>]. In efforts to maximize energy savings, cooperative adaptive cruise control (CACC) is utilized to maintain relatively short headways. However, when platooning is attempted in the real world, small transient accelerations caused by imperfect control result in observed energy savings being less than expected values.</div><div class="htmlview paragraph">This study analyzes the performance of a recently developed nonlinear model predictive control (NMPC) platooning strategy over challenging terrain. The NMPC strategy is compared to the previous proportional-integral-derivative (PID) control scheme in terms of headway, commanded torque, and fuel rate variances along with the total fuel consumed per lap. These comparisons reveal that the NMPC based controller’s ability to optimize headway variation while considering upcoming grade disturbances reduces the harshness of commanded torque and fuel rate transients. These platoon behavior changes result in significant fuel energy consumption reductions. In all platooning configurations analyzed, the NMPC strategy consumed less fuel than the comparable PID based data. This is best exemplified by findings from platoons with increased headway spacing. When compared to PID platoon control, the NMPC produced 25.5% and 31.6% fuel consumption decreases for the final truck in four-truck platoon configurations when targeting 50 foot and 100 foot follow distances, respectively. These results suggest that the NMPC implementation minimizes extraneous acceleration events associated with rigid PID headway adherence.</div></div>

In this work, the pH neutralization process is described by a neural network Wiener (NNW) model. A nonlinear Model Predictive Control (NMPC) is established for the considered process. The main difficulty that can be encountered in NMPC is solving the optimization problem at each sampling time to determine an optimal solution in finite time. The aim of this paper is the use of global optimization method to solve the NMPC minimization problem. Therefore, we propose in this work, to use the Self Organizing Migrating Algorithm (SOMA) to solve the presented optimization problem. This algorithm proves its efficiency to determine the optimal control sequence with a lower computation time. Then the NMPC is compared to adaptive PID controller, where we propose to use the SOMA algorithm to formulate the PID in order to determine the optimal parameters of the PID. The performances of the two controllers based on the SOMA algorithm are tested on the pH neutralization process.

In this paper, application of a modified nonlinear model predictive control (NMPC) method with simplified optimization problem has been studied. The NMPC scheme used here is based on Wiener model which requires only little more effort in development than a standard linear step-response model, yet offers superior characterization of systems with highly nonlinear gains. In this work, Wiener model has been incorporated into model predictive control scheme in a way that effectively removes the nonlinearity from the control problem. This scheme is based on identification of inverse nonlinear part of the Wiener model. The presented method results in an optimization problem with the implementation and computation levels of difficulty that are nearly equivalent to the case where the plant and constraints are linear. A standard quadratic programming (QP) algorithm is enough to solve the optimization problem. PH neutralization, as a highly nonlinear plant, is used in simulations to demonstrate the efficiency of the presented method.

This brief presents a nonlinear model predictive control (NMPC) method for the trajectory tracking problem of an autonomous underwater vehicle (AUV). By augmenting the desired output trajectory to a reference dynamical system, the tracking task can fit into the standard NMPC framework, which effectively avoids possible numerical difficulties in the following fast NMPC implementation. To relieve the conflict between short sampling period and high demand of online calculation, Ohtsuka's continuation/generalized minimal residual (C/GMRES) algorithm is investigated. In order to handle the realistic constraints on the AUV thrusters, we incorporate the log barrier functions into the cost function and modify the C/GMRES algorithm. Several different reference trajectories are tested using the identified dynamic model of the Saab SeaEye Falcon open-frame ROV/AUV, which demonstrate the effectiveness and efficiency of the proposed fast algorithm for the AUV tracking control.

The rapid acceleration of urbanization and the surge in car ownership necessitate efficient automatic parking solutions in constricted spaces to address the escalating urban parking issue. To optimize space utilization, enhance traffic efficiency, and mitigate accident risks, a method is proposed for smooth, comfortable, and adaptable automatic parking trajectory planning. This study initially employs a hybrid A* algorithm to generate a preliminary path, then fits the velocity and acceleration based on a cubic polynomial. The kinematic constraints of the vehicle and obstacle avoidance constraints are then meticulously defined, and a coupled nonlinear model predictive control (NMPC) method is employed to optimize the trajectory. Compared to the hybrid A* algorithm, the optimized trajectory demonstrates superior space utilization and improved smoothness. The experimental results indicate that the proposed method performs effectively in automated parking tasks in confined spaces, suggesting promising applications and broad prospects for future.