China
Africa
Explore chapters and articles related to this topic
Modeling and simulation in building automation systems
Published in Jan L.M. Hensen, Roberto Lamberts, Building Performance Simulation for Design and Operation, 2019
This section introduces a set of illustrative examples of least squares versus Bayesian model calibration as well as online and offline optimization approaches applied to commercial and residential buildings using model predictive control. MPC is a control methodology which seeks strategies through time that minimize an objective or cost function, based on the predictions of an energy model. In the context of building systems, MPC allows for discovery of near-optimal operation strategies that minimize, for example, the energy use, carbon dioxide emissions, or dollar cost of a facility. The first examples describes global temperature setpoint optimization for a large office building, the second example illustrates the development of optimal and near-optimal window opening strategies for a small mixed mode ventilated office building, and the third example describes building control strategies in consideration of energy use, energy expense, peak demand, economic demand response revenue, and frequency regulation revenue.
Motion Control Issues
Published in Bogdan M. Wilamowski, J. David Irwin, Control and Mechatronics, 2018
Roberto Oboe, Makoto Iwasaki, Toshiyuki Murakami, Seta Bogosyan
Recently, new approaches to prove the stability of a telerobotic equipment in a discrete-time frame-work, based on passivity, have been proposed in Secchi et al. (2003). Another promising approach is the bilateral generalized predictive control (BGPC) proposed in Slama et al. (2007), which is based on an extension of model predictive control (MPC). MPC is an advanced method for process control that has been used in several process industries such as chemical plants, oil refineries, and in robotics area. The major advantages of MPC are the possibility to handle constraints and the intrinsic ability to compensate large or poorly known time delays. The main idea of MPC is to rely on dynamic models of the process in order to predict the future process behavior on a receding horizon and, accordingly, to select command input with respect to the future reference behavior. Motivated by all the advantages of this method, the MPC was applied to teleoperation systems (Sheng and Spong 2004). The originality of the approach proposed in Slama et al. (2007) lies in an extension of the general MPC, so-called bilateral MPC (BMPC), allowing to take into account the case where the reference trajectory is not a priori known in advance due to the slave force feedback. The bilateral term is employed to specify the use of the signal feedback, which alters the reference system dynamic in the controller.
Adaptive Min-max Model Predictive Control for Field Vehicle Guidance in the Presence of Wheel Slip
Published in Dan Zhang, Bin Wei, Robotics and Mechatronics for Agriculture, 2017
Xu Wang, Javad Taghia, Jay Katupitiya
A very promising control method for achieving high precision path tracking is Model Predictive Control (MPC) due to its receding optimization and predictive ability. MPC has been successfully used in many industrial applications such as oil-refining and power systems (Qin and Badgwell, 2003; Richalet, 1993; Arnold and Andersson, 2011). In the recent past, researchers have shown an interest in applying MPC to path tracking. While there is an abundance of satisfactory research results, the majority of them use the assumption of pure rolling without sliding (Backman et al., 2009; Yaonan et al., 2010). As emphasized before, this assumption is invalid when it comes to the control of field vehicles in farming environments. Moreover, classical MPC is not inherently robust (Garcia et al., 1989), therefore it is necessary to design controllers taking the wheel slips into account. The work presented by Backman et al. (2010), took into account the wheel slip and used extended Kalman filter to compensate for the slippage, however, this approach is not robust due to the assumption of the Gaussian distribution of slip, which is not a reliable assumption. Lenain et al. (2005, 2006) used an extended kinematic model with two slip angles representing front and rear slip to design a control law and then created a sliding estimation algorithm to obtain the two slip angles. The results show acceptable performance, however, the noise levels on the two estimated slip angles were problematic.
Implementation of real-time model predictive heating control for a factory building using ANN-based lumped modelling approach
Published in Journal of Building Performance Simulation, 2023
Seon Jung Ra, Han Sol Shin, Cheol Soo Park
MPC is a control strategy that searches optimal control variable(s) to minimize(s) an objective function over a finite prediction horizon. Figure 7 shows a schematic of the MPC framework used for the target building. For MPC, it is important to determine hyperparameters (sampling period, prediction horizon, and control horizon) because the hyperparameters influence the overall performance of the system (Drgoňa et al. 2020). In this study, the sampling period was set to one minute. The prediction and control horizons were equally set to ten minutes in order to reflect the degree of a change in the indoor air temperature as shown in Table 3. In the factory building, the authors observed significant changes in the indoor temperatures at each zone over 10 min when the unit heaters were fully operated (Table 3). In this regard, the authors assumed that ‘one-step ahead’ prediction in this application would be appropriate.
Modelling and model predictive control for a bicycle-rider system
Published in Vehicle System Dynamics, 2018
When riding a bicycle, riders have the ability to predict how their control actions will affect the bicycle’s state based on the practical experiences of the system’s physical characteristics and dynamics. MPC with the control output is optimised based on the prediction of the system’s future behaviours and constraints. This provides a good solution for modelling the bicycle-rider control. The block diagram of an MPC is depicted in Figure 3. By using a prediction model, the MPC can calculate the system’s future outputs to a set of future inputs . The future inputs are generated via an optimisation process which intends to minimise the quadratic cost function of the predicted errors and the control efforts. Physical limitations of the system are considered as constraints in the optimisation process. In this study, the MPC-based rider model is used to perform roll-angle tracking tasks. Steering torque actions on the bicycle’s front fork and the leaning torque of the rider’s upper body are used as control inputs.
Real-time predictive control of HVAC systems for factory building using lightweight data-driven model
Published in Journal of Building Performance Simulation, 2023
With the development of communication systems including data storage, computing for analysis, and automatic controllers, MPC can be widely studied and adopted for building systems to overcome existing simple control methods. MPC is a control process wherein the optimization solver searches for optimal control variable(s) that minimizes the cost function for the prediction horizon using a model describing the system dynamics (Figure 10). MPC has many advantages, including the ability to handle multiple constraints and uncertainties, ability to handle time-varying system dynamics and a wide range of operating conditions, and the use of a cost function to achieve multiple targets (Afram and Janabi-Sharifi 2014).