Centralized Model Predictive Control Research Articles

A centralised model predictive control framework for logistics management of coordinated supply chains of perishable goods

ABSTRACT This paper proposes a centralised model predictive control framework to address logistics management of supply chains of perishable goods. Meeting customer specific requirements is decisive to gain a competitive advantage in supply chain management. This fact motivates stakeholders to address solutions that continuously improve supply chain operations. The solution proposed in this work considers the supply chain as a dynamical system in a state-space representation where different categories of commodities, namely common goods and perishable goods, are included. Additionally, the dynamical model is able to store information of the complete supply chain regarding the quantity of commodities and the due time associated to the perishable goods. A centralised controller then collects the supply chain state information and optimises the commodity flow based on the model prediction over a fixed time horizon. The model predictive control solution assigns just-in-time commodity flows, schedules production according to customer demand (pull system) and monitors work-in-progress and in-transit commodities. The success of the proposed control approach is demonstrated in a numerical simulation of a three-tier supply chain following three distinct management policies.

Computation of Least-Conservative State-Constraint Sets for Decentralized MPC With Dynamic and Constraint Coupling

We address the problem of synthesizing state-constraint sets for a fully decentralized Model Predictive Control (MPC) scheme. We consider linear time-invariant discrete time systems, with subsystems possibly coupled in both dynamics and state constraints. For each individual subsystem we employ a set-based framework to compute the state-constraint sets, that are used to synthesize local tube-based MPC controllers. The offline problem that computes the constraint sets explicitly ensures that the feasible regions of the MPC controllers are non-empty, and whenever the controllers are feasible, the overall system constraints are satisfied with the least conservativeness possible. We demonstrate the closed-loop performance of the decentralized scheme, assessed with respect to centralized MPC, using a numerical example.

Layer-to-Layer Predictive Control of Inkjet 3-D Printing

This article develops and experimentally validates a distributed predictive control algorithm for closed-loop control of inkjet 3-D printing to handle constraints, e.g., droplet volume bounds, as well as the large-scale nature of the 3-D printing problem. The large number of decision variables, i.e., droplet volumes at each grid point, in high resolution inkjet 3-D printing makes centralized methods extremely time-consuming, thus, a distributed implementation of the controller is necessary. First, a graph-based height evolution model that captures the liquid spreading dynamics is described. Based on this model, a scalable closed-loop control algorithm using distributed model predictive control (MPC) that can reduce computation time significantly is designed and experimentally implemented. The performance and efficiency of the algorithm are shown to outperform open-loop printing and closed-loop printing with existing centralized MPC methods.

Energy Management-Based Predictive Controller for a Smart Building Powered by Renewable Energy

This paper presents a smart building energy management system (BEMS), which is in charge of optimally controlling the sustainable operation of a building-integrated-microgrid (BIM). The main objective is to develop an advanced high-level centralized control approach-based model predictive control (MPC) considering variations of renewable sources and loads. A finite-horizon planning optimization problem is developed to control the operation of the BIM. The model can be implemented as a BEMS for the BIM to manipulate the indoor temperature and optimize the operation of the system’s units. A centralized MPC-based algorithm is implemented for the power management scheduling of all sub-systems as well as power exchanges with the electrical grid. The MPC algorithm is verified over case studies applied to two floors residential building considering the climate condition of a typical day of March, where the effects of both loads and thermal resistance of building shell on the operation of the BIM are analyzed via numerical simulations. The analysis shows that 96% of the total electrical load has been fulfilled by the local production where 23% represents the total electric output of the micro-CHP and 73% is the renewable energy production. The deficit, which represents only 4%, is purchased from the electrical distribution network (EDN).

Centralized and distributed Model Predictive Control for the maximization of the thermal power of solar parabolic-trough plants

This paper proposes a new centralized Model Predictive Control (MPC) algorithm for the maximization of the thermal power obtained with a parabolic-trough collector field. The optimal operation of the plant is achieved by controlling a set of valves located at the beginning of each loop of collectors, which allow to outperform the response achieved with traditional control approaches for parabolic-trough plants.Unfortunately, the computational complexity of the proposed MPC controller hinders its application in real-time for medium and large parabolic-trough power plants. Consequently, this paper also proposes a logic-based distributed Model Predictive Control algorithm, which approaches the performance of the centralized MPC but entailing a much lower computational load.The proposed controllers are tested by simulation using a model of the collector field ACUREX (Almería, Spain) along a 2-h synthetic DNI profile. The results obtained show that the proposed distributed algorithm is able to perform quite close to the centralized one. Moreover, the analysis of the numerical results (in terms of achieved power) shows that the use of valves at the beginning of each loop substantially improve the achieved thermal power, that the achieved performance using a local controller is significantly lower than using a global one, and that the maximization of the thermal power does not imply the maximization or minimization of the outlet temperature.

A centralized multi-objective model predictive control for a biventricular assist device: An in vitro evaluation

Control of a biventricular assist device (BiVAD) is more challenging than control of a left ventricular assist device due to the process interactions between control loops in a multi-input-multi-output system. Hence, a single centralized multi-objective model predictive controller (CMO-MPC) has been developed to control a BiVAD. The CMO-MPC aims to: 1) adapt pump flow rate according to the Frank-Starling mechanism, 2) avoid ventricular suction, and 3) avoid vascular congestion. The CMO-MPC was benchmarked against a constant-speed (CS) setting in exercise, postural change, and systemic vascular resistance change tests in a mock circulation loop. The CMO-MPC increased pump flow rate from 5.0 L/min to 7.6 L/min in the exercise scenario, which was higher than the pump flow rate in the CS setting (6.0 L/min). In the postural change test, right ventricular end diastolic pressure (RVEDP) decreased to a minimum at 0.1 mmHg and 2.0 mmHg in the CS setting and the CMO-MPC, respectively, indicating that the CMO-MPC could minimize the risk of ventricular suction (with higher minimum RVEDP than the CS setting) when there was a sudden decrease in venous return. In all tests, the CMO-MPC could adapt pump flow rate without resulting events of ventricular suction and vascular congestion.

A novel distributed model predictive control method based on a substructuring technique for smart tensegrity structure vibrations

This paper considers distributed model predictive control (DMPC) for the vibrations of smart tensegrity structures with input saturation. A distributed control strategy is necessary to enable the autonomy in the individual subsystems, reduce the computational burden of centralized control, and enhance the fault tolerance of control systems. In this work, the dynamic model for smart tensegrity structures equipped with active piezoelectric actuators is generated first by the finite element method. Then, based on the substructuring technique, the entire tensegrity structure system is decomposed into a series of multilevel subsystems. For each subsystem, the continuous-time predictive model equations are discretized by the explicit form of the Newmark-β algorithm. Finally, based on the parametric variational principle, the DMPC problem with input saturation is transformed into a series of linear complementarity problems. The proposed method provides a simple, unified and flexible multilevel distributed control framework for solving the vibration control problems of smart tensegrity structure systems. Numerical experiments illustrate that the proposed DMPC method is effective and exhibits performance comparable to that of a typical centralized model predictive control (CMPC) method. In addition, the constraints of the control input saturation for the proposed DMPC method can be satisfied.

Distributed Model Predictive Control for Cooperative Landing

Abstract We design, implement and test two control algorithms for autonomously landing a drone on an autonomous boat. The first algorithm uses distributed model predictive control (DMPC), while the second combines a cascade controller with DMPC. The algorithms are implemented on a real drone, while the boat’s motion is simulated, and their performance is compared to a centralized model predictive controller. Field experiments are performed, where all algorithms show an ability to land while avoiding violation of the safety constraints. The two distributed algorithms further show the ability to use longer prediction horizons than the centralized model predictive controller, especially in the cascade case, and also demonstrate improved robustness towards breaks in communication.

Model Predictive Control for the Coordination of Autonomous Vehicles at Intersections

Abstract The paper describes a method for defining the crossing order of autonomous vehicles at a non-signalized intersection. The safe passage of vehicles through the intersection is achieved by a centralized Model Predictive Control (MPC) method. The minimization of traveling time or energy consumption and the condition of collision-avoidance are incorporated in the control design. The goal of the research is to guarantee safe passage of the autonomous vehicles by taking uncertainties of position measurements into consideration. The operation of the proposed control method is demonstrated by simulation examples made in a high-accuracy simulation environment to present its efficiency.

Effective Continuous-flow Supply Chains Using Centralized Model Predictive Control

This paper proposes three different formulations of a centralized Model Predictive Control framework to manage the logistics of continuous-flow Supply Chains subject to fluctuating demand. The Supply Chain is modeled as a dynamical system composed of several players handling commodities from the production phase to the retail phase. Additionally, commodities are categorized according to their characteristics. An external control agent continuously gathers information regarding Supply Chain operation. Using that information, the control agent monitors the inventory of the retailer and assigns the commodity quantity to replenish it, adopting a Model Predictive Control algorithm. Three different formulations of the Model Predictive Control algorithm are designed based on the inventory of the retailer: i) constant inventory, ii) dynamical heuristic inventory, and iii) dynamical control inventory. These formulations are simulated for a Supply Chain operating under a just-in-timemanagement policy.

Guest Editorial: Medpower 2018 Selected Papers

Guest Editorial: Medpower 2018 Selected Papers

Distributed Model Predictive Control of Steam/Water Loop in Large Scale Ships

In modern steam power plants, the ever-increasing complexity requires great reliability and flexibility of the control system. Hence, in this paper, the feasibility of a distributed model predictive control (DiMPC) strategy with an extended prediction self-adaptive control (EPSAC) framework is studied, in which the multiple controllers allow each sub-loop to have its own requirement flexibility. Meanwhile, the model predictive control can guarantee a good performance for the system with constraints. The performance is compared against a decentralized model predictive control (DeMPC) and a centralized model predictive control (CMPC). In order to improve the computing speed, a multiple objective model predictive control (MOMPC) is proposed. For the stability of the control system, the convergence of the DiMPC is discussed. Simulation tests are performed on the five different sub-loops of steam/water loop. The results indicate that the DiMPC may achieve similar performance as CMPC while outperforming the DeMPC method.

A Distributed Predictive Control of Energy Resources in Radiant Floor Buildings

This paper studies the impact of using different types of energy storages integrated with a heat pump for energy efficiency in radiant-floor buildings. In particular, the performance of the building energy resources management system is improved through the application of distributed model predictive control (DMPC) to better anticipate the effects of disturbances and real-time pricing together with following the modular structure of the system under control. To this end, the load side and heating system are decoupled through a three-element mixing valve, which enforces a fixed water flow rate in the building pipelines. Hence, the building temperature control is executed by a linear model predictive control, which in turn is able to exchange the building information with the heating system controller. On the contrary, there is a variable action of the mixing valve, which enforces a variable circulated water flow rate within the tank. In this case, the optimization problem is more complex than in literature due to the variable circulation water flow rate within the tank layers, which gives rise to a nonlinear model. Therefore, an adaptive linear model predictive control is designed for the heating system to deal with the system nonlinearity trough a successive linearization method around the current operating point. A battery is also installed as a further storage, in addition to the thermal energy storage, in order to have the option between the charging and discharging of both storages based on the electricity price tariff and the building and thermal energy storage inertia. A qualitative comparative analysis has been also carried out with a rule-based heuristic logic and a centralized model predictive control (CMPC) algorithm. Finally, the proposed control algorithm has been experimentally validated in a well-equipped smart grid research laboratory belonging to the ERIGrid Research Infrastructure, funded by European Union's Horizon 2020 Research and Innovation Programme.

Developing a centralized automatic control system to increase flexibility of water delivery within predictable and unpredictable irrigation water demands

Irrigation districts modernization has become a key strategy to address unreliable water resources and improve water use efficiency in irrigated regions. However, improving water use efficiency will be achieved when off-farm water distribution systems are integrated with on-farm practices within irrigation districts. In the present study a modernization alternative for irrigation canal operation, a centralized automated control system, was designed to handle both of known and unknown patterns of demanded water dictating by different irrigation systems in the farms. This alternative was developed in response to on-farm water management transitions from traditional, surface irrigation methods, to modern irrigation systems including the pressurized systems. The capability of Centralized Model Predictive Control (CMPC) was investigated for dealing with a set of complex water distribution test scenarios including predictable and unpredictable ±50% demand variations through the irrigation canal. Altus canal of the Lugert-Altus Irrigation District in southwest Oklahoma, USA is selected as the test case due to the recent challenges occurred in water distribution management in the Altus canal. The transition from traditional irrigation methods (i.e., furrow irrigation) to modern pressurized systems of subsurface drip irrigation (SDI) systems in large regions of the district caused the mentioned challenges. The results showed that when the demands were predictable, the controller successfully regulated water levels so that no off-take within the entire canal suffered from inadequate water delivery. Although the changes in the demand were considerable, stable canal operation was provided using the CMPC for this scenario. The designed controller also had an outstanding performance in dealing with extreme, unpredictable irrigation water taking test scenarios. The controller performed well with a quick hydraulic response to the changes, especially in the downstream canal reaches. The developed system started storing extra water within the canal reaches and employed it as a local inline reservoir to compensate delay times of water traveling within the irrigation canal. The stored water within the canal was entirely used after a sudden increase in any unknown demands.

Centralised model‐predictive decoupled active–reactive power control for three‐level neutral point clamped photovoltaic inverter with preference selective index‐based objective prioritisation

This study presents a single-stage grid-tied three-level neutral point clamped photovoltaic inverter with a centralised model-predictive decoupled active–reactive power control. The proposed centralised model predictive control (CMPC) incorporates the constraints of maximum power extraction, dc-link capacitor voltage balancing and active–reactive power tracking in a single objective function. The dc-link voltage of the inverter is regulated to its reference for extracting the maximum power. In order to eliminate the impact of reactive power exchange on floating dc-link voltage regulation, a decoupled active–reactive power control is used in the CMPC. Furthermore, a preference selective index-based dynamic weighting factor selection approach is introduced to maintain the relative importance between the power tracking and dc-link capacitor voltage balancing. The proposed control approach eliminates the outer dc-link voltage control loop and also, the empirical approach required for the selection of weighting factors. As a result, it ensures an optimal control action in each sampling period to improve the steady-state and dynamic tracking performance of the control objectives. The proposed control approach is experimentally verified by using a 1.2 kW laboratory-scale prototype and the results are presented to demonstrate its effectiveness compared to the classical proportional–integral-based model predictive control.

Economic model predictive control of combined thermal and electric residential building energy systems

This article investigates the potential of economic model predictive control of complex residential energy systems with electric coupling to the public grid. The examined system includes a battery energy storage system, photovoltaic power generation, an air-to-water heat pump, thermal energy storage and a building model. The said power generation provides energy for electric loads as well as domestic hot water and space heating. Model predictive control algorithms manage the energy system by nonlinear global optimization. Within this optimization, a time-varying state space model, which is derived from the energy system simulation model, reflects the system dynamics. Owing to the resulting high complexity, two algorithms for distributed model predictive control are developed. In addition, the developed approaches are compared to a common reference control concept as well as centralized model predictive control. For the comparison of annual operational costs, current German energy prices and subsidies are implemented into the economic calculation. Results show an improved performance of the developed approaches with 11.6% cost reduction in comparison to the reference. This is achieved through an increase of the heat pump seasonal performance factor by 3.4%, reduced curtailment of electrical photovoltaic energy to 21.5% of the reference value and prevention of auxiliary heater operation. Furthermore, increased photovoltaic self-consumption by the heat pump results in a slight reduction of battery storage operation. In conjunction with monetary and energetic advantages, model predictive control increases the comfort in regards to the violation of minimum limits for the comfort criteria.

Flexible operation of coal fired power plant integrated with post combustion CO2 capture using model predictive control

The growing demand for CO2 capture from coal-fired power plant (CFPP) has increased the need to improve the dynamic operability of the integrated power generation-CO2 capture plant. Nevertheless, high-level operation of the entire system is difficult to achieve due to the strong interactions between the CFPP and post combustion CO2 capture (PCC) unit. In addition, the control tasks of power generation and CO2 removal are in conflict, since the operation of both processes requires consuming large amount of steam. For these reasons, this paper develops a model for the integrated CFPP-PCC process and analyzes the dynamic relationships for the key variables within the integrated system. Based on the investigation, a centralized model predictive controller is developed to unify the power generation and PCC processes together, involving the key variables of the two systems and the interactions between them. Three operating modes are then studied for the predictive control system with different focuses on the overall system operation; power generation demand tracking and satisfying the CO2 capture requirement. The predictive controller can achieve a flexible operation of the integrated CFPP- PCC system and fully exert its functions in power generation and CO2 reduction.

A Hierarchical MPC Scheme for Coordination of Independent Systems With Shared Resources and Plug-and-Play Capabilities

This paper describes a hierarchical control scheme for the coordination of independent, asymptotically stable, and square systems with joint input and output constraints. At the higher layer of the control structure, a centralized model predictive control (MPC) problem is solved at a slow rate based on a reduced order model of the overall system to fulfill the global output request under shared resources. At the lower layer, fast MPC controllers are designed for each subsystem to remove the effects of the model mismatch introduced at the higher layer, satisfy local constraints and optimize the individual performance. The proposed algorithm allows to dynamically modify the control configuration according to plug-and-play operations. Theoretical results of feasibility and convergence are proven and a simulation study is reported to witness the potentialities of the method.

Distributed Building Energy Storage Units for Frequency Control Service in Power Systems

Abstract This paper presents a dynamical Building-to-Grid (BtG) framework with explicit hierarchical interactions between Transmission System Operator (TSO), Distribution System Operator (DSO), and Building energy storage units (BLD). Such a proposed hierarchical multi-level framework, yields a realistic model of the future power system, which can be used to present the interaction between different stakeholders in the power system. Using a centralized model predictive control (MPC) to optimize BLD, DSO, and TSO control variables simultaneously over a common time horizon, we demonstrate the potential role of individual buildings in frequency control of the power system, by introducing electrical energy storage units in the BtG framework. We present the power grid regulative capacity of building storage flexibility by coupling BLD, DSO, and TSO decisions and solving the corresponding finite-horizon optimal control problem in a receding horizon fashion. An extended IEEE five bus power system case study coupled with two DSOs, each contains seven buildings, is presented to illustrate functionality of the developed integrated hierarchical BtG model and the flexibility obtained via building energy storage units for the frequency control service in power system.

Distributed economic model predictive control of wastewater treatment plants

In this work, we consider the distributed economic model predictive control (EMPC) of a wastewater treatment plant described by Benchmark Simulation Model No. 1 and compare its performance with two commonly used control methods. The plant is decomposed into two subsystems based on the topology of the process, the requirement in our distributed control system design and the computational considerations. A local EMPC controller is designed for each subsystem. Two different designs are considered. In the first design, the centralized model is used in each subsystem EMPC controller design; and in the second design, a subsystem model is used in each subsystem EMPC design. The performance of these two distributed EMPC designs are compared with a centralized model predictive control (MPC) scheme and a centralized EMPC scheme from different aspects including effluent quality, operating cost, and computational efficiency. The simulation results demonstrate that the distributed EMPC with each subsystem EMPC controller designed based on the entire system model is more favourable in terms of control performance.