Predictive Control System Research Articles

Real-world validation of safe reinforcement learning, model predictive control and decision tree-based home energy management systems

Real-world validation of safe reinforcement learning, model predictive control and decision tree-based home energy management systems

Data-centric predictive control with tuna swarm optimization-backpropagation neural networks for enhanced wind turbine performance

Wind energy is a significant renewable resource, but its efficient harnessing requires advanced control systems. This study presents a Data-Centric Predictive Control (DPC) system, enhanced by a Tuna Swarm Optimization-Backpropagation Neural Network (TSO-BPNN) for predictive wind turbine control. It's like a smart tool that uses innovative fusion of deep learning, predictive Control, and reinforcement learning. Unlike traditional control methods, the proposed approach uses real-time data to optimize turbine performance in response to fluctuating wind conditions.The system is validated using simulations on the FAST platform, which demonstrate its superior performance in two critical operational regions. Specifically, in Region II, where the objective is to maximize power extraction from the wind, the DPC achieves a 1.07 % reduction in overshoot and an improvement of 36.14 units in steady-state error compared to traditional methods. The response time remains comparable to existing Model Predictive Control (MPC) strategies, ensuring real-time applicability without sacrificing efficiency. In Region III, where maintaining constant power output is crucial, the DPC outperforms both the baseline and MPC methods, reducing overshoot by 0.58 % and improving accuracy by 17.27 units compared to the baseline method. These results highlight the effectiveness of the proposed DPC system in optimizing turbine performance under variable wind conditions, offering a significant improvement over traditional methods in both accuracy and control precision.

Cascade Model Predictive Control for Enhancing UAV Quadcopter Stability and Energy Efficiency in Wind Turbulent Mangrove Forest Environment

Unmanned aerial vehicle (UAV) modelling and control, particularly in quadrotors with high position-orientation coupling, present significant challenges for practical applications, such as environmental monitoring missions in windy mangrove forests. Conventional control strategies like the PID controller, often employed in simulations due to their simplicity, often underperform in real-world scenarios due to their linear assumptions. This research proposes a novel hierarchical cascaded model predictive control system for altitude, attitude, and battery efficiency for quadrotor in mangrove area. This control system addresses computational complexity by decomposing the overall MPC strategy into two distinct schemes, one for translational displacements and another for rotational movements, enhancing the UAV's resilience to wind turbulence, a significant disturbance factor in mangrove environments. Rigorous simulation and experiment test flight involving complex trajectory tracking and windy conditions demonstrate the proposed controller's superior performance compared to conventional PID controller, particularly in terms of stability, disturbance rejection, underscoring its potential for UAV applications in challenging environments.

Design of the predictive management and control system for combined propulsion complex

The object of this researching is the process of maneuvering a sea-based vehicle under compressed conditions, which requires one hundred percent reserve of thrusters (THRs) of various modifications and locations. The main problem is the provision of energy-efficient control over the ship's motion at low speed in the horizontal plane using a high-level predictive controller. The hierarchy of the motion control system (MCS) is usually divided into several levels with the help of a high-level motion controller and the THR motor control distribution algorithm. This allows for a modular software structure where a high-level controller (HLC) can be designed without using comprehensive information about the THR motors. However, for a certain reference of THR configurations, such a decoupling can lead to reduced control performance due to the limitations of HLC regarding the physical constraints of the vessel and the behavior of MCS. The main results of the researching are methods to improve control performance using a nonlinear model predictive control (MPC) as a basis for the designed motion controllers due to its optimized solution and ability to consider constraints. A decoupled system was implemented for two simple motor tasks showing dissociation problems. The shortcomings were eliminated through the development of a nonlinear MPC controller, which combines the motion controller and the distribution of control over THR motors. To preserve the discrete modularity of the control system and achieve adequate performance, a nonlinear MPC controller with time-varying constraints was designed. This has made it possible to take into account the current limitations of the THR control system, increase the accuracy of control, and reduce the response time of the system by 10 %.

Co-simulation Architecture and Platform Establishment Method for Cloud-Based Predictive Cruise Control System

Co-simulation Architecture and Platform Establishment Method for Cloud-Based Predictive Cruise Control System

Assessment of impact load effect on self-sensing of cement composites incorporating hybrid silicon carbide-graphite under various environmental conditions

This study investigates the self-sensing abilities of hybrid silicon carbide-graphite cement composites (HSCGC) under various environmental conditions, focusing on dynamic impacts. Integrating silicon carbide (SiC) and graphite enhances the mechanical and electrical properties of these composites. Tests showed that composites with 30 % SiC and 2 % graphite had a 14.1 % increase in compressive strength and up to 85 % decrease in electrical resistivity. Water absorption reduced by 3.25 %, indicating better durability. Impact resistance ratios (IRR) varied with humidity and impact angles, with the highest IRR observed at 70 % humidity and a 30-degree angle. Self-sensing responses indicated that higher SiC and graphite content maintained stable conductivity, especially at 60 and 90-degree angles. Temperature changes affected resistivity, negligible at −10°C and increased at 45°C. The study concludes that HSCGCs are suitable for smart construction materials, capable of real-time environmental and structural monitoring, advancing predictive models and control systems.

Digital-Twin Predictive Control of Nonlinear Systems With Time Delays, Unknown Dynamics, and Communication Delays.

With the advancement of computing technology and big data technology, digital twins have gradually been applied in various fields, such as manufacturing, energy, and healthcare. This article studies the predictive control of nonlinear dynamic systems using digital twins. Based on a digital-twin control system framework, predictive control is discussed for three different nonlinear systems with time delays: 1) known nonlinear systems; 2) unknown nonlinear systems; and 3) unknown nonlinear cyber-physical systems. Both a digital-twin predictive control strategy and a digital-twin control predictor are proposed to compensate for time delays and communication delays actively. With the strategy and predictor, the digital-twin controller of a time-delay nonlinear system can be designed to achieve the desired performance based on the nonlinear system without time delays, which vastly simplifies the controller design procedure. A digital model is constructed using data to deal with unknown nonlinear dynamics. The three different closed-loop digital-twin predictive control systems are analyzed to derive a unified stability criterion. The simulation results show how the proposed digital-twin predictive control method performs well for nonlinear systems with time delays, unknown dynamics, and/or communication delays.

A Data-Driven Online Prediction Model for Battery Charging Efficiency Accounting for Entropic Heat

This study proposes a charging efficiency calculation model based on an equivalent internal resistance framework. A data-driven neural network model is developed to predict the charging efficiency of lithium titanate (LTO) batteries for 5% state of charge (SOC) segments under various charging conditions. By considering the impact of entropy change on the open-circuit voltage (OCV) during the charging process, the accuracy of energy efficiency calculations is improved. Incorporating battery data under various charging conditions, and comparing the predictive accuracy and computational complexity of different hyperparameter configurations, we establish a backpropagation neural network model designed for implementation in embedded systems. The model predicts the energy efficiency of subsequent 5% SOC segments based on the current SOC and operating conditions. The results indicate that the model achieves a prediction error of only 0.29% under unknown charging conditions while also facilitating the deployment of the neural network model in embedded systems. In future applications, the relevant predictive data can be transmitted in real time to the cooling system for thermal generation forecasting and predictive control of battery systems, thereby enhancing temperature control precision and improving cooling system efficiency.

Research on random dynamic cylinder deactivation control strategy of engine based on nonlinear model prediction

In the process of cycle cylinder deactivation, the different proportion of cylinder deactivation between adjacent working cycles will cause the problem of uneven operation of each cylinder. In this paper, a random dynamic cylinder deactivation (RD-CDA) control method based on nonlinear model prediction is proposed. The RBF neural network model of the RD-CDA system is built. The K-means algorithm is applied to the offline training of the center position c of the hidden layer node. The P-nearest algorithm is used to train the hidden layer node width σ offline. The recursive least squares (RLS) algorithm is applied to train the output layer weight vector w online. The online adaptive change of the model is realized, and the model verification is carried out. Then, the nonlinear model predictive control system structures with the minimum torque fluctuation and brake specific fuel consumption rate (BSFC) are built respectively. Under different working conditions, the average torque fluctuation per cycle is improved by about 9% to 18.75%, and the average BSFC is improved by about 2 % -3.9 %, which verifies the effectiveness of this strategy in reducing the dynamic fluctuation of random dynamic cylinder deactivation torque and improving fuel economy.

Analyzing the implementation of predictive control systems and application of stored data in non-residential buildings

In non-residential buildings, building energy management systems (BEMS) and the application of data hold significant promise in reducing energy consumption. Nevertheless, BEMS have different levels of complexity, benefit, and limitation. Despite the advanced technologies and improvements in building operation, there is a clear gap in the actual performance of buildings that has been attributed to the adoption of advanced technologies. Consequently, there is an increasing need for researchers and practitioners to study current practices in order to identify and address the challenges that compromise the core objectives of BEMS. For this reason, this paper aims to validate three research questions: (i) to examine the current state of BEMS and its functionalities; (ii) to analyze the type of control used; (iii) and to determine the availability of historical data compiled by BEMS and its application in non-residential buildings. A survey of 676 buildings and interviews with building professionals were conducted. The findings confirmed that most of the buildings applied BEMS with scheduled control. In addition, a lack of digitized data for analysis and predictions was detected. Indeed, only 0.60% of the investigated buildings implemented predictive control. Finally, using hierarchical clustering analysis, responses were grouped to analyze similarities between them. The study findings help to develop targeted actions for implementing predictive control in non-residential buildings.

Research on energy consumption optimization of predictive cruise control considering the state of the leading vehicle

To comprehensively assess the economic performance of electric vehicles (EVs) cruise planning and mitigate conflicts with leading vehicles, a dynamic programming-based predictive cruise speed optimization control system was developed. This system comprises two layers: an upper V2X-based LSTM-DP cruise speed planning system and a lower adaptive cruise control system based on sliding mode control (SMC), and the mode switching strategy of the speed and distance tracking controller is designed. Initially, a segmented uniform variable motion algorithm processed the historical driving data of the leading vehicle, significantly enhancing the LSTM neural network's acceleration prediction capability, particularly on inclines, with an improvement in accuracy of 37.8 %. Subsequently, an adaptive cruise tracking controller was implemented to track the planned vehicle speed. Simulation tests on actual roads demonstrated that the LSTM-DP approach not only ensures driving safety but also markedly improves fuel efficiency over conventional constant speed cruising methods when planning is effective. Moreover, even in scenarios where planning fails, the LSTM-DP strategy can still achieve energy savings of up to 37 % without compromising the safety distance between vehicles. Finally, hardware-in-the-loop (HIL) testing confirmed the system's efficacy under real-time operational conditions.

Header Height Detection and Terrain-Adaptive Control Strategy Using Area Array LiDAR

During the operation of combine harvesters, the cutting platform height is typically controlled using manual valve hydraulic systems, which can result in issues such as delays in adjustment and high labor intensity, affecting both the quality and efficiency of the operation. There is an urgent need to enhance the automation level. Conventional methods frequently employ single-point measurements and lack extensive area coverage, which means their results do not fully represent the terrain’s variations in the area and are prone to local anomalies. Given the inherently undulating terrain of farmland during harvesting, a control strategy that does not adjust for minor undulations but only for significant ones proves to be more rational. To this end, a sine wave superposition model was established to simulate three-dimensional ground elevation changes, and an area array LiDAR was used to collect 8 × 8 data for the header height. The effects of mounds and stubble on the measurement results were analyzed, and a dynamic process simulation model for the solenoid valve core was developed to analyze the on/off delay characteristics of a three-position four-way electromagnetic directional valve. Moreover, a physical model of the hydraulic system was constructed based on the Simscape module in Simulink, and the Bang Bang switch predictive control system based on position threshold was introduced to achieve early switching of the electromagnetic directional valve circuit. In addition, an automatic control system for cutting platform height was designed based on an STM32 microcontroller. The control system was tested on the hydraulic automatic control test rig developed by Shanxi Agricultural University. The simulation and experimental results demonstrated that the control system and strategy were robust to output disturbances, effectively enhancing the intelligence and environmental adaptability of agricultural machinery operations.

Constrained model predictive control for networked jump systems under denial‐of‐service attacks and time delays

AbstractThis article addresses the problem of model predictive control of networked jump systems in the presence of DoS attacks and time delays. In the structural framework of the network predictive control system, we mathematically model the networked jump system by using Markov chains to describe the time delays and a polytope model to describe the jump system phenomenon, considering the properties of DoS attacks and time delays. Based on this, we propose a strategy to lessen the effect of network constraints on the control performance of the system. This strategy involves the corresponding control inputs from the control sequence for real‐time active compensation. It includes adjusting the control sequence application length variation based on the duration of the DoS attacks and time delays at each moment. In addition, we demonstrate the recursive feasibility of the control strategy and the global asymptotic stability of the control system from a theoretical perspective through the Lyapunov stability theory. Finally, the effectiveness of the proposed strategy is verified by simulation arithmetic.

Deep learning model predictive control of a high-density polyethylene reactor with a physics-guided sequence-to-sequence model with memory

In chemical process industry, the input-output data of processes display complex nonlinear dynamics and strong influences of unobserved hidden states. Their behavior must be modeled using nonlinear time series with an observer-predictor structure. To address this, a sequence-to-sequence model with a memory layer was proposed for a high-density polyethylene slurry reactor. The memory layer retains the chronological contribution of the observer and predictor and effectively captures the lengthy time response. A physics-guided approach was adopted to ensure the directional consistency between input and output variables in key control loops. In this way, a deep learning model can be obtained with historical data and there is no need for plant tests. The resulting model can be used in a nonlinear model predictive control system that not only quickly navigates different grade transitions but also provides steady-state control even though key internal state variables such as catalyst activity change.

Optimization of the Parameters of a Model Predictive Control System for an Industrial Fractionator

The problem of parametric synthesis of a model predictive control (MPC) system by the chemical process of production of the kerosene fraction of an industrial fractionator under conditions of constraints and uncertainty is considered. The optimal parameters of the MPC algorithm are obtained as a result of solving the problem of multi-criteria optimization, taking into account the intervally specified parameters of the plant model.

Optimization of the Parameters of a Model Predictive Control System for an Industrial Fractionator

Optimization of the Parameters of a Model Predictive Control System for an Industrial Fractionator

Research on accurate morphology predictive control of CFETR multi-purpose overload robot

The CFETR multipurpose overload robot (CMOR) is a critical component of the fusion reactor remote handling system. To accurately calculate and visualize the structural deformation and stress characteristics of the CMOR motion process, this paper first establishes a CMOR kinematic model to analyze the unfolding and working process in the vacuum chamber. Then, the dynamic model of CMOR is established using the Lagrangian method, and the rigid-flexible coupling modeling of CMOR links and joints is achieved using the finite element method and the linear spring damping equivalent model. The co-simulation results of the CMOR rigid-flexible coupled model show that when the end load is 2000 kg, the extreme value of the end-effector position error is more than 0.12 m, and the maximum stress value is 1.85 × 108 Pa. To utilize the stress-strain data of CMOR, this paper designs a CMOR morphology prediction control system based on Unity software. Implanting CMOR finite element analysis data into the Unity environment, researchers can monitor the stress strain generated by different motion trajectories of the CMOR robotic arm in the control system. It provides a platform for subsequent research on CMOR error compensation and extreme operation warnings.

Design of Predictive Control System for Lane Change in Autonomous Vehicle

The automobile business is introducing a lot of autonomous vehicles in the modern day. Lane changes are one of the most complicated urban scenarios in which autonomous vehicles are used. Self-driving automobiles must thus interact with human-driven vehicles in a certain way. In this work, we concentrate on the autonomous vehicle's lane-changing control system for obstacle avoidance. This study employs a predictive control system as its methodology. The vehicle's next movements can be predicted by this control system. The vehicle's position, which is adjusted by the steering angle, is the controllable variable. The vehicle's position, which is adjusted by the steering angle, is the controllable variable. It is clear from the numerical simulation results that the predictive control system executes control actions on lane changes correctly, avoiding collisions with the running vehicle obstacles. RMSE (Root-Mean Square Error) is a performance metric that is derived from the difference between the vehicle's lateral position and the reference trajectory value. The RMSE of the planned predictive control is 0.9681.

Development of a fed-batch-scale anaerobic co-digestion control system based on multivariable output error state space and model predictive control

This paper describes a feeding control system for fed-batch-scale anaerobic co-digestion. The proposed system is based on the multivariable output error state space and uses model predictive control to regulate the biogas flow output of anaerobic co-digestion. This control system combines predictive modeling, receding-horizon optimization, and state feedback. An accurate linear model of anaerobic co-digestion is obtained through the multivariable output error state space method using experimental data, and the accuracy of the model is evaluated using a goodness-of-fit index and the root mean square error. After augmenting the model, a predictive control system is established. The system states are observed through a Kalman filter, and the constrained receding-horizon optimization problem is solved using quadratic programming. Finally, the stability of the control system is demonstrated from both open- and closed-loop perspectives. The control system is then constructed and used in actual experiments, with the target flow output set at 100 NmL/h. The final flow output indicates that the control system achieves good control performance. This research enables the biogas output of anaerobic co-digestion to be controlled from a linear perspective through a concise control algorithm with strong practical application prospects.

Agent-Based System for Continuous Control and its Application to Activated Sludge Process

This paper presents a concept of architecture and ontology layouts for development of multiagent model-based predictive control systems. The presented architecture provides guidelines to simplify development of agent-based systems and improve its maintainability. The proposed multiagent system (MAS) layout is split into multiple subsystems that include agents dedicated to performing assigned tasks. MAS implementation was prepared which can use provided algorithms, actuators and can react to changes in its environment to reach the best available control quality. An example of MAS based on the proposed architecture is shown in application of dissolved oxygen (DO) concentration control in a laboratory activated sludge setup with biological reactor. For that application, MAS incorporates agent-based controllers from Boundary-Based Predictive Controllers (BBPC) family. Presented experiments prove flexibility, resilience and online reconfiguration ability of the proposed multiagent system.