Model-based Predictive Control Controller Research Articles

Methodology for energy management strategies design based on predictive control techniques for smart grids

This article focuses on the development of a general energy management system (EMS) design methodology using on model-based predictive control (MPC) for the control and management of microgrids. Different MPC-based EMS for microgrids have been defined in the literature; however, there is a lack of generality in the proposed that would facilitate adapting to new architectures, energy storage system technology, nature of the bus, application, or purpose. To fill this gap, a novel general formulation that is parameterizable, simple, easily interpretable, and reproducible in different microgrid architectures is presented. This is the result of the development of a novel methodology, which is also presented. It considers the state space formulation of the controller from the initial modelling phase, from the dynamics of the energy storage systems represented by their models to the subsequent definition of the optimisation problem. This is developed through the design of the general cost function and the formulation of constrains, by means of general guidelines and reference values. To evaluate the performance of the developed methodology, simulation tests were carried out for four different microgrid architectures, with different applications and objectives, also considering different generation conditions, demand profiles, and initial conditions. The results showed that, with some simple guidelines and regardless of the case study, the developed MPC controller design methodology can address the technical-economic optimisation problem associated with energy management in microgrids in an easy and intuitive way.

Ultraprecise Controller for Piezoelectric Actuators Based on Deep Learning and Model Predictive Control.

Piezoelectric actuators (PEA) are high-precision devices used in applications requiring micrometric displacements. However, PEAs present non-linearity phenomena that introduce drawbacks at high precision applications. One of these phenomena is hysteresis, which considerably reduces their performance. The introduction of appropriate control strategies may improve the accuracy of the PEAs. This paper presents a high precision control scheme to be used at PEAs based on the model-based predictive control (MPC) scheme. In this work, the model used to feed the MPC controller has been achieved by means of artificial neural networks (ANN). This approach simplifies the obtaining of the model, since the achievement of a precise mathematical model that reproduces the dynamics of the PEA is a complex task. The presented approach has been embedded over the dSPACE control platform and has been tested over a commercial PEA, supplied by Thorlabs, conducting experiments to demonstrate improvements of the MPC. In addition, the results of the MPC controller have been compared with a proportional-integral-derivative (PID) controller. The experimental results show that the MPC control strategy achieves higher accuracy at high precision PEA applications such as tracking periodic reference signals and sudden reference change.

Procedures to Design Multi-Model-Based Predictive Controller Applied to BLDC Drive

One of the main reasons that prevent the use of model-based predictive control (MPC) in power electronics and motor drive is the MPC design. The processes related to these fields have fast dynamics, and most of the MPC design guides involve the control of slow dynamic processes, which allow high sampling times and, consequently, a large time for the control action calculation. Moreover, practically only the linear model-based predictive controller can have small computational cost, which causes difficulty for MPC application in nonlinear processes as electrical motors. Therefore, this paper presents a MPC design theory for first-order control models, as the brushless direct current (BLDC) motor drive, assuming direct speed control (without current inner loop). Thus, the MPC cost function tune is obtained according to prediction model parameters, the study of optimization gain curves and, optionally, an extended closed-loop root locus analysis. To deal with the motor nonlinearities, both multi-model approach and six-step modulation are employed. This paper also discusses current peak transients, salient pole BLDC motor modeling, integral action included in the MPC control action and solutions for Hall effect sensor nonlinear speed measurement. Simulation and experimental results confirm the proposed design theory as well as the other discussed issues.

Potential of Model-Free Control for Demand-Side Management Considering Real-Time Pricing

This paper presents a detailed description of data predictive control (DPC) applied to a demand-side energy management system. Different from traditional model-based predictive control (MPC) algorithms, this approach introduces two model-free algorithms of artificial neural network (ANN) and random forest (RF) to make control strategy predictions on system operation, while avoiding the huge cost and effort associated with learning a grey/white box model of the physical system. The operating characteristics of electrical appliances, system energy consumption, and users’ comfort zones are also considered in the selected energy management system based on a real-time electricity pricing system. Case studies consisting of two scenarios (0% and 15% electricity price fluctuation) are delivered to demonstrate the effectiveness of the proposed approach. Simulation results demonstrate that the DPC controller based on ANN pays only 0.18% additional bill cost to maintain users’ comfort zones and system economy standardization while using only 0.096% optimization time cost compared with the MPC controller.

Objective function with economic cost in MPC

The Model Based Predictive Control (MPC) is an advanced control technique that can deal with of multivariable systems with interactions and considerable dead times, nonlinearities and restrictions on the variables. Different types of objective functions can be used in the MPC control algorithm with specific parameter settings for each type of objective function. In this work a MPC control strategy for distillation column using an objective function with economic cost was implemented. The MPC controller with the objective function proposed worked properly, stabilizing the distillation column when it has undergone a change in the setpoint and also a series of disturbances in the feed. Using energy point of view, the implemented controller is more efficient than the standard MPC controller.

MOPSO-based predictive control strategy for efficient operation of sensorless vector-controlled fuel cell electric vehicle induction motor drives

This paper introduces an optimal control strategy of model-based predictive control (MPC) based on multiobjective particle swarm optimization (MOPSO) for a sensorless vector control induction motor, which is used in a fuel cell electric vehicle drive system. The proposed MPC-MOPSO algorithm is implemented to tune the weighting parameters of the MPC controller to tackle all the conflicting objective functions. The paper handles the following fitness functions: minimizing the speed error, minimizing the torque ripple, minimizing the DC-link voltage ripple, and minimizing machine flux ripple. Computer simulations studies have been completed utilizing MATLAB/Simulink with a specific end goal of assessing the dynamic performance of the proposed MPC-MOPSO optimal controller and comparing it with single-objective particle swarm optimization and traditional PI controllers. The simulation results demonstrate the good dynamic response of the proposed MPC-MOPSO optimal tuning strategy over the traditional PI controllers for more accurate tracking performance through the whole speed range, especially at starting conditions and load change disturbances.

Predictive zone control of pressure management for water supply network systems

In this paper we address the problem of pressure management in water supply system (WSS) network. The model-based predictive control (MPC) strategies have some important features to deal with WSS. By hydraulic analysis of WSS, the predictive model is derived from the dynamic model and static model of WSS. Through WSS, the consumers' demands are required to be met at all times according to some operational constraints that must be satisfied. The constraints of flow through actuators, the water level of reservoirs and the consumer areas' pressure demand are determined by a specific system. In this work, we develop a constrained MPC controller that considers the zone control of the pressure outputs and incorporates steady state economic targets in the control cost function. The designed management strategies are applied to a case study and simulation results, covering different aspects, are provided. The output nodal pressure can be controlled in the desired zone by optimal scheduling the actuators of the WSS. If the variation range of reservoir's water level is broader, the rate of flow through the actuators is gentle, and vice versa.

A MPC approach for optimal generation scheduling in CSP plants

Thermal energy storage (TES) systems allow concentrated solar power (CSP) producers to participate in a day-ahead market. Therefore, the optimal power scheduling problem can be posed, whose objective is the maximization of profits derived from electricity sales. The daily generation schedule has to be offered in advance, usually the previous day before a certain time, thus an electricity price and weather forecast must be carried out. This paper proposes a model-based predictive control (MPC) approach for optimal scheduling in CSP plants. This approach has a dual purpose: (1) the periodic update of the generation schedule to track the schedule that has been committed to by means of the most recent electricity price and weather forecast and information about the plant state and (2) the generation of the optimal schedule for the next day. As these two tasks are related, they are performed simultaneously. Therefore, the MPC sliding window is composed of a first time interval to track the committed schedule and a second time interval to generate the next schedule for the following hours. This is then offered as the generation schedule for the next day at the appropriate time. The proposed approach is applied, in a simulation context, to a 50MW parabolic trough collector-based CSP plant with molten-salt-based TES. The chosen criterion to track the committed schedule is the even distribution of the possible generation error within the first interval. A case-study with overestimated initial DNI forecast is undertaken. The results show that the MPC control with short-term DNI forecast significantly improves the above-mentioned objective and allows for a reduction of the deviation from the scheduled generation, when compared with the case without short-term DNI forecast.

Model-Based Power Plant Master Control

In this paper a coordinated master control for a solid fuel power plant has been developed and the performance evaluated in terms of tracking capability, stability and robustness. The control strategy has been model-based predictive control (MPC) and it was evaluated on a nonlinear process model of the Vattenfall power plant Idbäcken in Nyköping, Sweden. The developed master MPC with gain scheduling has better performance compared to the existing PID controller which has been thoroughly studied and tuned in a previous project. The robustness of the proposed master MPC controller against common disturbances and parameter variation has been investigated and it shows that the proposed controller is more robust than the existing PID controller.

An investigation of predictive control for aluminum wheel casting via a virtual process model

An advanced control solution, specifically Model-based predictive control (MPC), has been developed for a low-pressure die casting (LPDC) process used to manufacture aluminum alloy wheels. The control solution was developed offline using a high-fidelity 3D finite element model as a virtual process on which open and closed loop trials were performed. Open loop trials were used to assess the influence of process variables and to facilitate the development of a reduced-order state-space model to predict the approximate input–output behaviour of the casting process. An MPC controller employing this state-space model was used to regulate die temperatures in the virtual process during disturbance scenarios. Two disturbances typically found in the industrial process were simulated, that is, the temperature behaviour of the molten metal in the industrial process and the length of time the die remains open after the cast wheel is ejected from the dies. These disturbances cause the die temperatures to deviate from their optimal values, which can result in defective wheels. The developed control solution improved the process performance by rejecting the simulated disturbances and maintaining the die temperatures near their optimal values.

LATERAL GUIDANCE OF HEAVY-DUTY VEHICLE PLATOONS USING MODEL-BASED PREDICTIVE CONTROL

Heavy-duty vehicle platoons are a possible future form of organising goods traffic. Only the leading vehicle is manually driven while all the following vehicles are automatically controlled in longitudinal and lateral direction. This paper presents a concept for the generation of reference variables as well as for a model-based predictive control (MPC) scheme to be used for the lateral control within the path tracking. It enables the following vehicles to track the path of the leading vehicle directly by using on-board sensors as well as the existing communication structure of the platoon. It is shown that the presented control concept is feasible and it is evaluated in simulations of tractor-semitrailer platoons using Matlab/Simulink. The results demonstrate the effectiveness of the reference generation as well as of the proposed MPC control scheme. The application to real vehicles will be part of the current project KONVOI at RWTH Aachen.

Performance of Model-based Predictive Control of the Ventilation Rate with Axial Fans

A new algorithm for ventilation rate control in mechanically ventilated buildings was developed and evaluated on a laboratory test installation. It was demonstrated that the current problems with traditional control of axial fans, such as instability due to dynamic wind actions, could be overcome by using a model-based predictive control (MBPC) algorithm. In the laboratory test installation, the performance of the MBPC algorithm was compared to the performance of a traditional proportional, integral, derivative (PID) controller in an airflow range from 1000 to 5000 m3/h (20–100% of fan capacity) and in a normal pressure range from 0 to 60 Pa. The accuracy and the stability of the ventilation rate control and the energy consumption were studied for both control algorithms. It was concluded that the main advantage of a MBPC algorithm over a PID controller lies in its stability in a wider range of the ventilation rates and pressure differences due to an automatic adaptation of the control parameters. The PID algorithm with fixed settings did not generate a stable control of the ventilation rate when the target value deviated more than 500 m3/h (16% of centre capacity) from the ventilation rate that it was optimised for, whereas the MBPC controller remained stable throughout the complete working range of the fan.

Implementation of Predictive Model Based Control on Programmable Logic Controller

Model-based predictive control (MBPC) is one of the first and most popular non-standard control methods. Although MBPC has around 30 year history, it was usually implemented in industrial control systems using PC computers. Technological progress made control equipment more powerful last years. Modem programmable logic controllers (PLCs) are now capable to run more computation consuming algorithms. The paper presents an application of one of the commercially used MBPC control methods - Predictive Functional Control (PFC). The standard first-order nonlinear system with polynomial approximation is proposed as internal model. The application was made for flow control of the fluid in the installation of chemical reactor. Presented application shows that nonlinear control can be necessary even at a very low level in the control hierarchy.