Model For Control Purposes Research Articles

A physically interpretable data-driven surrogate model for wake steering

Abstract. Wake steering models for control purposes are typically based on analytical wake descriptions tuned to match experimental or numerical data. This study explores whether a data-driven surrogate model with a high degree of physical interpretation can accurately describe the redirected wake. A linear model trained with large-eddy-simulation data estimates wake parameters such as deficit, center location and curliness from measurable inflow and turbine variables. These wake parameters are then used to generate vertical cross-sections of the wake at desired downstream locations. In a validation considering eight boundary layers ranging from neutral to stable conditions, the far wake's trajectory, curl and available power are accurately estimated. A significant improvement in accuracy is shown in a benchmark study against two analytical wake models, especially under derated operating conditions and stable atmospheric stratifications. Even though the results are not directly generalizable to all atmospheric conditions, locations or turbine types, the outcome of this study is encouraging.

Safety model construction for a complex automatic transportation system

The Aim of the paper is to consider approaches to the analysis of a safety model of complex multi-loop transportation systems comprising not completely supervised subsystems. Method. For the description of a safety model, the paper uses systems theoretic process analysis (STPA) methods and the principles specified in ISO/PAS 21448:2019 (SOTIF). Result. The paper shows drawbacks of the FTA and FMEA local risk analysis methods and demonstrates a demand for some universal approach based on the combination of STPA and control theory. It gives an overview of the major stages of such analysis for the safety model of complex transportation systems exemplified by the Moscow Central Circle, which provide a feedback for safety evaluation of a transport control system under development. The paper analyzes the feasibility of using a virtual model for control purposes in the form of a so-called “supervised artificial neural network”.Conclusion. Today, railways are actively testing autonomous systems (with no driver onboard) that apply as their subsystems automatic perception modules using machine learning. The introduction of the latter into the control loop complicates the task of hazard analysis and safety evaluation of such systems using conventional FTA and FMEA methods. The construction of a safety model of such complex multi-loop transportation systems comprising not completely supervised subsystems that use machine learning methods with not completely predictable behavior requires the application of a systems approach to the analysis of unsafe scenarios along with the compilation of a scenario library and the formalization of a hazard model’s description, pertaining to the boundaries of various control loops as well, in order to reduce the regions of unknown unsafe scenarios for autonomous transportation systems under development.

Cyberphisical adaptive manufacturing control systems

Cyberphysical systems of adaptive manufacturing control are considered. It has been shown that the concept of Cyberphysical systems can be used for both the construction and analysis of adaptive systems. A generalized technology of research is proposed. A generalized model has been formed both for manufacturing in general and for structured manufacturing with a universal structure. All possible perturbations in the system and their compensating controls are considered. It is noted that the prior art publications use simpler system structures. It has been shown that the proposed model can be used in the study of newly studied systems with different structures, including those not previously discussed. Initially, we consider the technology of building a model that allows you to ensure the optimal mode and the prompt transition to the production of new products. Emphasis is placed on obtaining numerical data. A specific technology for using the model for control purposes is then proposed.

OpenHPL for Modelling the Trollheim Hydropower Plant

An increase of tightly integrated, renewable energy sources with highly varying production leads to a greater need for flexible hydropower plants to “balance” the intermittent production from these sources. From a systems perspective, the intermittency aspect constitutes a disturbance, while the tight integration leads to a multivariable systems, both of which causes increased challenges for good control. This instigates the need for a model based analysis and synthesis tool which can describe dynamic phenomena, supports multiphysics problems, and allows for immediate use in advanced control analysis and design. Previously, an open source Hydro Power Library (OpenHPL) have been developed within the multiphysics Modelica eco-system which satisfies the above requirements: OpenHPL contains a set of model units relevant for hydropower systems, as well as examples of tools for analysis of these models for control purposes. Thus far, OpenHPL has been validated by comparison against other simulation tools. However, a real test of OpenHPL is to use it to describe a real hydropower plant using minimal plant information, to tune model parameters against experimental data, and to validate the model. As a case study, experimental data from the Trollheim hydropower plant in Norway have been made available, and used to test OpenHPL. By flow sheeting a hydropower model in OpenModelica using OpenHPL, the model can be simulated from within a script language (Python via the OMPython API, Julia via the OMJulia API) and further analyzed using analysis tools in the script language. Based on default model parameter suggestions from OpenHPL, least squares model fitting was carried out in Python, and the model was validated with good model fit. Important parts of the modelling phase were the development of a new friction model for the Francis turbine, and iterations on the description of the turbine outlet geometry. The results are complemented with a discussion of possible uses of the model.

Forecasting District Heating Demand using Machine Learning Algorithms

Short-term forecasting of thermal energy demand is critical to optimally manage on-site renewable energy generation and the charge and discharge of energy storage devices in district heating and cooling (DHC) systems. As part of a larger study on advanced predictive control for a solar district heating system with 52 homes – the Drake Landing Solar Community (DLSC) – this paper investigates the use of Machine Learning algorithms to predict the aggregated heating load of the community.In this study, the initial approach to estimate the heating load of the DLSC employed a piecewise linear regression based on the outdoor air temperature. Such an approach yields significant errors, in particular when weather forecasts are used instead of actual outdoor air temperature measurements. It has been found that Machine Learning algorithms, such as decision trees, can significantly improve the accuracy of predicted heating loads by incorporating the effect of additional influencing factors (e.g., time of the day, day of the week, solar radiation, etc.). In this study, the predicted heating demand obtained from different algorithms are compared under two different scenarios; (a) by using actual weather conditions from measured data; (b) by using weather forecasts. The potential implementation of such models for control purposes is discussed.

A simplified model for friction compensation in precise feed drives

Friction is a complex phenomenon with negative effects on the precision of positioning systems. In this work, a single-state friction model for control purposes, aimed to simplify the implementation of control algorithms while keeping a good performance in friction compensation, is analysed. In particular, the model is capable or representing the Stribeck curve, rate independent presliding hysteresis with non-local memory and non-drifting behaviour. This model has been tested on an experimental setup based on a linear motor, and its performance is compared to that of the generalised Maxwell-slip model, showing better performance on the tracking error.

Nonlinear Modeling and Estimation Techniques for a Fed-Batch Prototype Bioprocess

This paper presents nonlinear modeling and estimation techniques dedicated to a fed-batch prototype bioprocess. More precisely, this prototype bioprocess is an aerobic microbial growth process coupled with an enzyme-catalyzed reaction, which is widely used in bioindustry. As a modeling procedure, the present work proposes an innovative methodology, namely the bond graph approach. In order to obtain a useful model for control purposes, it is necessary to derive estimation strategies for unmeasurable states and kinetic parameters. By using some model structural properties, an asymptotic state observer is derived. Also, the nonlinear and uncertain kinetic rates of this process are on-line estimated by using a high-gain observer. The performance and behavior of estimation techniques are analyzed via numerical simulations.

Solution for Inverse Kinematics of 5 DOF Serial Link MCM Machine

The paper presents the forward and inverse kinematics problems solution for 5 DoF automated serial link MCM cutting and welding machine. Particularly, the solution was developed for specific application to be implemented in industry. The developing of the mathematical model for control purposes, implementing algorithms for trajectory finding and mathematical model for forward and inverse kinematics solution are described in the work. The solution is presented as simulation in RobWork Studio software [, as well as a 3D interactive animation implemented in specifically developed software environment. The result of given solution are evaluated and directions of further work concerning the implementation of obtained solution in the actual industrial application are defined.

Multivariable feedback linearizing control of Chlamydomonas reinhardtii photoautotrophic growth process in a torus photobioreactor

The objective of this paper is to design and to validate a nonlinear multivariable controller, based on the exact feedback linearization technique, which has the capacity to stabilize the photoautotrophic microalgae growth in photobioreactor regardless of the operation point or the transient trajectory. Two measurable outputs were selected, namely the biomass concentration and the pH, whose simultaneous control can be realized by manipulating the dilution rate and the injected carbon dioxide gas flowrate. A mechanistic dynamical model describing the interactions between physicochemical and biological phenomena inside the photobioreactor was used as a starting point for the controller design. Given its complexity, the model was reduced by differentiating the states with slow dynamics from the ones with fast dynamics that were converted into algebraic expressions. In addition, a dynamic time-varying expression was derived for the concentration of hydrogen ions (negative antilogarithm of pH) thus obtaining an appropriate I/O model for control purposes. The degree of interaction between I/O channels was determined through a relative gain array analysis, in order to establish if the system allows the implementation of decentralized SISO controllers or requires a centralized MIMO controller. The feedback linearization method was associated with a model based technique to furnish the immeasurable variables required by the nonlinear control algorithm. The nonlinear controller was implemented on a laboratory torus photobioreactor piloted at constant incident light flux intensity throughout the photoautotrophic growth of a Chlamydomonas reinhardtii culture, its efficiency being proved under several operating points.

Linear Parameter-Varying Modeling of Electric Vehicle Air Conditioning System

This article presents the approach of quasi LPV (Linear Parameter-Varying) modeling techniques for an air conditioning system of an electric vehicle. Vehicle air conditioning systems are strongly non-linear systems and it is a challenging task to get a precise real time model for control purposes. Therefore, an LPV method is first introduced to estimate the air conditioning system. Experimental results show that the LPV model delivers a very high accuracy for the COP (Coefficient Of Performance) estimation, that can’t be reached by traditional identification methods. Some discussion about the model structure and its application are presented and a non-linear LPV model structure similar to the Hammerstein structure is proposed.

Intelligent Control Structure for Control of Integrated Wastewater Systems

The paper proposes an approach to design a structure for control of integrated wastewater treatment plant-sewer system under full range of disturbance inputs. The control activities are hierarchically structured into a multilevel-multilayer form. A newly developed robust Model Predictive Control method for uncertain and output constrained dynamic systems is a key control technology employed at Optimising Control Level. This level is further decomposed into control layers operating at different time scales: slow, medium and fast. Appropriate models for control purposes is described Needed data are delivered by measuring system and Extended Kaiman Filter. A Supervisory Control Level is located at a top of the control hierarchy.

In-line Particle Size Measurement for Control of Jet Milling

Jet milling is suitable for the production of very fine powders with a well-defined particle size distribution (PSD). The major disadvantage of the process is the poor energy efficiency. In practice jet mills have to be operated below their maximum capacity in order to avoid the production of off-specification material, due to operation disturbances. Therefore, it is important to automate operation with a real-time PSD instrument for quality control. In this paper, a study for in-line particle size measurement using laser diffraction is presented to improve operation control. This measurement system uses a novel CMOS pixel array, which has a robust and flexible feature for the application. It also includes an in-line cell system, where direct measurement is implemented in the product particle stream coming from the mill and an air purge for the cell window allows continued operation. Experiments were conducted on a 10-inch spiral jet mill pilot plant with alpha-Al2O3 as test material. The milling process is investigated by varying the operational conditions. The measurement results are useful to build a grinding model for control purpose.

Experience of using the neural network approach for identification of MSF desalination plants

Mathematical models of physical and engineering systems are frequently of high dimension, possessing interacting dynamic phenomena. Information processing requirements for experimenting with these models for control purposes are excessive. It is therefore natural to techniques that reduce the computational effort. In this paper, the neural network approach for nonlinear identification of the multistage flash desalination plant (large scale system) is presented through the manipulation of collected input-output data from the plant. Results of utilizing such an approach are also discussed.

A heuristic for multi-project scheduling with limited resources in the housing industry

A heuristic scheduling and control model is developed for the multi-project, resource constrained scheduling environment. The model is developed from actual firm data using the Kurtulus and Davis approach to select appropriate heuristic decision rules. Resource profiles, calculation of multi-project resource constrained slack and the schedule are outputs of the model. A routine to update current projects status is provided with this model for control purposes in an ongoing operational environment. While developed for the housing industry, the modelling approach has broad application in a number of multi-project environments.