Grey-box Model Research Articles

Nonlinear and linearized gray box models of direct-write printing dynamics

PurposeThe purpose of this study is to develop nonlinear and linearized models of DW printing dynamics that capture the complexity of DW while remaining integrable into control schemes. Control of material metering in extrusion-based additive manufacturing modalities, such as positive displacement direct-write (DW), is critical for manufacturing accuracy. However, in DW, transient flows are poorly controlled due to capacitive pressure dynamics – pressure is stored and slowly released over time from the build material and other compliant system elements, adversely impacting flow rate start-ups and stops. Thus far, modeling of these dynamics has ranged from simplistic, potentially omitting key contributors to the observed phenomena, to highly complex, making usage in control schemes difficult.Design/methodology/approachThe authors present nonlinear and linearized models that seek to both capture the capacitive and nonlinear resistive fluid elements of DW systems and to pose them as ordinary differential equations for integration into control schemes. The authors validate the theoretical study with experimental flow rate and material measurements across a range of extrusion nozzle sizes and materials. The authors explore the contribution of the system and build material bulk modulus to these dynamics.FindingsThe authors show that all tested models accurately describe the measured dynamics, facilitating ease of integration into future control systems. Additionally, the authors show that system bulk modulus may be substantially reduced through appropriate system design. However, the remaining build material bulk modulus is sufficient to require feedback control for accurate material delivery.Originality/valueThis study presents new nonlinear and linear models for DW printing dynamics. The authors show that linear models are sufficient to describe the dynamics, with small errors between nonlinear and linear models. The authors demonstrate control is necessary for accurate material delivery in DW.

Construction of a Valid Domain for a Hybrid Model and Its Application to Dynamic Optimization with Controlled Exploration

A hybrid model, also called a gray-box model, utilizes a black-box and white-box model simultaneously to capture the behaviors of a complex system. When available data are not sufficient to estimat...

Inclusive Multiple Models (IMM) for predicting groundwater levels and treating heterogeneity

An explicit model management framework is introduced for predictive Groundwater Levels (GWL), particularly suitable to Observation Wells (OWs) with sparse and possibly heterogeneous data. The framework implements Multiple Models (MM) under the architecture of organising them at levels, as follows: (i) Level 0: treat heterogeneity in the data, e.g. Self-Organised Mapping (SOM) to classify the OWs; and decide on model structure, e.g. formulate a grey box model to predict GWLs. (ii) Level 1: construct MMs, e.g. two Fuzzy Logic (FL) and one Neurofuzzy (NF) models. (iii) Level 2: formulate strategies to combine the MM at Level 1, for which the paper uses Artificial Neural Networks (Strategy 1) and simple averaging (Strategy 2). Whilst the above model management strategy is novel, a critical view is presented, according to which modelling practices are: Inclusive Multiple Modelling (IMM) practices contrasted with existing practices, branded by the paper as Exclusionary Multiple Modelling (EMM). Scientific thinking over IMMs is captured as a framework with four dimensions: Model Reuse (MR), Hierarchical Recursion (HR), Elastic Learning Environment (ELE) and Goal Orientation (GO) and these together make the acronym of RHEO. Therefore, IMM-RHEO is piloted in the aquifer of Tabriz Plain with sparse and possibly heterogeneous data. The results provide some evidence that (i) IMM at two levels improves on the accuracy of individual models; and (ii) model combinations in IMM practices bring ‘model-learning’ into fashion for learning with the goal to explain baseline conditions and impacts of subsequent management changes.

An investigation of knowledge protection practices in inter-organisational collaboration: protecting specialised engineering knowledge with a practice based on grey-box modelling

PurposeThis paper aims to report an interview study investigating knowledge protection practices in a collaborative research and innovation project centred around the semi-conductor industry. The authors explore which and how knowledge protection practices are applied and zoom in on a particular one to investigate the perspective of three stakeholders which collaborate: the SUPPLIER of a specialised machine, the APPLIER of this machine and a SCHOLAR who collaborates with both, in an effort to develop a grey-box model of the machine and its operation.Design/methodology/approachA total of 33 interviews have been conducted in two rounds: 30 interviews explore knowledge protection practices applied across a large project. Qualitative content analysis is applied to determine practices not well covered by the research community. A total of three follow-up interviews inspect one specific collaboration case of three partners. Quotes from all interviews are used to illustrate the participants’ viewpoints and motivation.FindingsSCHOLAR and APPLIER communicate using a data-centric knowledge protection practice, in that concrete parameter values are sensitive and hidden by communicating data within a wider parameter range. This practice balances the benefit that all three stakeholders have from communicating about specifics of machine design and operations. The grey-box model combines engineering knowledge of both SUPPLIER and APPLIER.Practical implicationsThe line of thought described in this study is applicable to comparable collaboration constellations of a SUPPLIER of a machine, an APPLIER of a machine and a SCHOLAR who analyses and draws insights out of data.Originality/valueThe paper fills a research gap by reporting on applied knowledge protection practices and characterising a data-centric knowledge protection practice around a grey-box model.

Development of Inverse Greybox Model-Based Virtual Meters for Air Handling Units

Energy submetering at the equipment level provides a tool to identify energy use anomalies and detect operational inefficiencies. While physical meters can be costly and difficult to install, virtual meters (VMs) overcome practical issues associated with physical meters and provide insights into critical unmeasured quantities. This article introduces an inverse greybox model-based virtual metering method to estimate the energy in an air handling unit (AHU). Models that represent the components typically found in AHUs are formulated using a data set from a highly instrumented AHU and combined into an integrated greybox model. The use of the integrated model to create VMs is demonstrated by using a data set from an independent AHU located in a large office building in Ottawa, ON, Canada. Model parameters are estimated by using the genetic algorithm and used in creating VMs that can estimate the heat supplied/extracted at the AHU level. In addition, the model is used to estimate a monthly average outdoor air fraction used by the AHU. The potential of the component models and VMs to detect operational inefficiencies and support operational decisions is demonstrated through illustrative examples. Note to Practitioners-This article presents a novel virtual metering algorithm to estimate the heating and cooling energy at the air handling unit (AHU) level. This virtual metering algorithm fills a gap in the literature and provides a tool that will help detect and interpret energy use anomalies, identify operational inefficiencies, and guide on-going commissioning of building energy systems. Facility managers, operators, and other stakeholders can use the insights gained from virtual metering to improve building operational performance. Future planned research includes developing virtual meters to characterize energy flows at the zone level and visualization methods to make inverse modeling results more accessible to different building stakeholders.

Fuzzy model predictive control for small-scale biomass combustion furnaces

This work presents a fuzzy model predictive controller for small-scale grate furnaces based on a newly derived biomass combustion model. Several local linear controllers are designed for a selected number of operating points utilizing a gap metric. The resulting local predictive controllers are merged with membership functions to form a global nonlinear fuzzy control structure. The presented framework intends to improve the transient and steady state operation by applying an optimal control strategy with state estimation and to cover the entire operating range of the furnace. The open loop results of the introduced combustion model are parameterized and cross-validated with measured data from a test furnace. In order to find suitable parameters for the grey-box model, a local sensitivity analysis is conducted to contribute to an efficient parameter estimation process. Closed loop simulation results of the fuzzy model predictive controller, a linear model predictive controller and a PI control algorithm are presented and compared. Based on the performance of the proposed fuzzy controller, its application, advantages and disadvantages are discussed. Additionally, the impact of the different controllers on the formation of carbon monoxide is investigated based on estimation models from literature. The simulation results show that the fuzzy model predictive controller performs best in the considered categories.

Closing the gap between simulation and measured energy use in home archetypes

Recent climate action plans set ambitious targets for large-scale home energy retrofit. These retrofit targets also recommend switching from oil boilers to electrical heat pumps. One area-based approach to retrofit is neighbourhood scale. Ideally with local community participation, neighbourhood retrofit improves a group of structurally similar homes, of the same archetype. Defined home archetypes already exist for 20 European countries thanks to previous research. That same research attempted steady-state calculations of space heating energy using the ISO 13790 Seasonal Method. Those calculations diverged from measured energy use by a large gap. Taking example home archetypes from Ireland, this study narrowed this energy-use gap by incorporating realistic parameters and schedules. Physical white-box models were constructed in DesignBuilder, and simulated by EnergyPlus with local weather data. The simulations combined realistic heating schedules, internal gains, infiltration and ventilation; converging simulation energy use with typical measurements and mostly achieved model calibration indices. Therefore, the resultant energy-use data was capable of tuning future grey-box models to investigate home energy use across ranges of heating patterns and heating setpoints. The aims of this study are threefold. First, devise a repeatable method to produce heat energy-use values by simulating white-box models of home archetypes, that enable tuning of future grey-box models. Second, specify the heating variables that result in acceptable energy-use gaps between simulation of home archetype models and their typical measurements. Third, test if the retrofits of three Irish home archetypes under a typical occupant schedule, achieve national climate action targets.

Application of a grey-box modelling approach for the online monitoring of batch production in the chemical industry

Abstract Model-based solutions for monitoring and control of chemical batch processes have been of interest in research for many decades. However, unlike in continuous processes, in which model-based tools such as Model Predictive Control (MPC) have become a standard in the industry, the reported use of models for batch processes, either for monitoring or control, is rather scarce. This limited use is attributed partly to the inherent complexity of the batch processes (e. g., dynamic, nonlinear, multipurpose) and partly to the lack of appropriate commercial tools in the past. In recent years, algorithms and commercial tools for model-based monitoring and control of batch processes have become more mature and in the era of Industry 4.0 and digitalization they are slowly but steadily gaining more interest in real-word batch applications. This contribution provides a practical example in this application field. Specifically, the use of a grey-box modeling approach, in which a multiway Projection to Latent Structure (PLS) model is combined with a first-principles model, to monitor the evolution of a batch polymerization process and predict in real-time the final batch quality is reported. The modeling approach is described, and the experimental results obtained from an industrial batch laboratory reactor are presented.

Gray-box virtual sensor of the supply air temperature of air handling units

Building automation system uses several networks of sensors for continuous monitoring of building control systems for energy-efficient operation. Physical sensors are costly and need frequent calibration. The accurate measurement of supply air temperature from the air handling units (AHUs) has an important effect on the control of cooling coil, supply air temperature and supply airflow rate delivered to rooms. In the case when such a sensor gives erroneous measurements, a virtual sensor can replace temporarily the faulty sensor, and it can also be used for automated fault detection of HVAC systems. This article proposes two different gray-box (models A and B) for predicting the supply air temperature of air handling units that were developed and tested using the measurements from two buildings. The models require the measurement of three variables (mixed air temperature, cooling coil valve signal, and chilled water inlet temperature). The results of both models are discussed and compared. A sliding window approach for cross-validation of the models was also carried out. The developed gray-box models A and B could be integrated into BAS for virtual measurement, virtual calibration, and fault detection in HVAC systems.

Black Box Modelling and Simulating the Dynamic Indoor Relative Humidity of a Laboratory Using Autoregressive–moving-average (ARMA) Model

Modelling and simulation of the dynamic indoor relative humidity behaviour of a building is essential to test any proposed thermal comfort controller and strategy in the building. Like other plants, the dynamic indoor relative humidity behaviour of a building can be developed based on the white box model, black box model, and grey box model. This research focuses on the usage of autoregressive–moving-average (ARMA) model, a type of black box model to represent the dynamic indoor relative humidity behaviour of Industrial Instrumentation Laboratory at Malaysia-Japan International Institute of Technology (MJIIT), UniversitiTeknologi Malaysia (UTM) Kuala Lumpur and uses the real recorded data from the laboratory and minimal knowledge regarding the physical characteristics of the humidity behaviour in the laboratory. The performance of the ARMA model developed in this research is compared with the real recorded data from the laboratory. Result obtained shows that the ARMA model is enough for modelling and simulating the dynamic indoor relative humidity behaviour of the laboratory

Hybrid-Learning Type-2 Takagi–Sugeno–Kang Fuzzy Systems for Temperature Estimation in Hot-Rolling

Entry temperature estimation is a major concern for finishing mill set-up in hot strip mills. Variations in the incoming bar conditions, frequent product changes and measurement uncertainties may cause erroneous estimation, and hence, an incorrect mill set-up causing a faulty bar head-end. In earlier works, several varieties of neuro-fuzzy systems have been tested due to their adaptation capabilities. In order to test the combination of the simplicity offered by Takagi–Sugeno–Kang systems (also known as Sugeno systems) and the modeling power of type-2 fuzzy, in this work, hybrid-learning type-2 Sugeno fuzzy systems are evaluated and compared with the results presented earlier. Systems with both empirically and fuzzy c-means-generated rules as well as purely fuzzy systems and grey-box models are tested. Experimental data were collected from a real-life mill; datasets for rule-generation, training, and validation were randomly drawn. Two of the grey-box models presented here reach 100% of bars with 20 °C or less prediction error, while two of the purely fuzzy systems improved performance with respect to purely fuzzy systems presented elsewhere, however it was only a slight improvement.

Distributed agent-based building grey-box model identification

The paper deals with an identification (calibration) of a building thermal model as a crucial part of modern control algorithms. It describes a modelling issue and parameter identification approach. There is presented an identification approach using a so called dual decomposition method which decomposes a large optimization problem into smaller local ones which are then solved by local agents. The local models are found using a grey-box calibration and by a coordination of agents’ mutual shared parameters a global consistency is obtained.The proposed method is tested on two examples; one is basically trivial, in the latter one, more complex, a model of a building simulated in the EnergyPlus program is obtained and offers promising results for (predictive) control-oriented purposes.

Identification of multi-zone grey-box building models for use in model predictive control

ABSTRACT Predictive controllers can greatly improve the performance of energy systems in buildings. An important challenge of these controllers is the need of a building model accurate and simple enough for optimization. Grey-box modelling stands as a popular approach, but the identification of reliable grey-box models is hampered by the complexity of the parameter estimation process, specifically for multi-zone models. Hence, single-zone models are commonly used, limiting the performance and applicability of the predictive controller. This paper investigates the feasibility of the identification of multi-zone grey-box building models and the benefits of using these models in predictive control. For this purpose, the parameter estimation process is split by individual zones to obtain an educated initial guess. A virtual test case from the BOPTEST framework is contemplated to assess the simulation and control performance. The results show the relevance of modelling thermal interactions between zones in the multi-zone building.

Thermal modelling of three-way mixing valves using Bézier curves for parameter estimation applications

Each mixing valve is characterized by its kvs-value and the valve characteristics. In mounted condition, the valve characteristics are deformed and the degree of the deformation is determined by the valve authority. Considering already installed mixing valves, neither the valve authority nor the course of the undeformed valve characteristics are exactly known. The lack of information poses a challenge when physically modelling the relationship between the inlet temperatures, the control signal and the outlet temperature.Based on Bézier curves, a universal non-linear grey-box model to describe the thermal valve behaviour is proposed. It is compared with a state space as well as a Hammerstein model and shows the advantages over black-box approaches. The trained model is very robust, able to predict measurement data very accurately and can be used in control design. The parameter estimation is performed using the Open Source platform JModelica.org.

Reduced-order models for assessing demand response with heat pumps – Insights from the German energy system

Heat pumps are regarded as one promising option for Demand Response (DR) in renewable-dominated energy systems. By activating the structural thermal mass of buildings, power consumption of heat pumps can be decoupled from heat demand. Assessing the system-wide DR potential from heat pumps requires adequate models to account for the thermal response of the building stock to flexible operation. In this study, reduced-order thermal response models of twelve representative German building types in three insulation states are provided based on grey-box modeling. The identified models reveal a good compromise between accuracy of the simulated indoor temperature (RMSE of on average 0.6 °C) and computational cost (acceleration by factor 250) compared to the complex building simulation software TRNSYS. Furthermore, these reduced-order models are employed in a case study to estimate the technical DR potential of heat pumps by passive storage at specified ambient conditions. Extrapolated to the future stock of German single family houses, the maximum shiftable heat pump load adds up to 57 GWel. The structural thermal mass of the building stock enables maximal 300 GWhel of electrical energy to be discharged at a very cold winter day and 185 GWhel to be charged at milder temperatures.

Design, fabrication, modeling and control of a fabric-based spherical robotic arm

In this paper we present a spherical soft robotic arm made from fabric. The inflatable arm has a small mass and is pneumatically actuated. A configuration is employed with only three actuators controlling the two rotational degrees of freedom of a spherical joint. This differs from the commonly employed antagonistic pairs, including four actuators for two degrees of freedom. The fabrication procedure of the lightweight and compliant system is discussed in detail and uses commonly available materials and tools. The capability of the robotic arm to adjust the joint stiffness as a function of the actuator pressures is investigated and characterized for different deflection directions. The static mapping from the actuator pressures to the orientation of the robotic arm is identified from data and the inverse mapping is employed in a position controller. The modeling and controller derivation are performed for three different stiffness levels demonstrating the ability of the spherical robotic arm to change the joint stiffness independently of controlling the position. The position tracking performance is experimentally evaluated by tracking a square trajectory. A comparison of the tracking performance for the different stiffness levels shows that accurate tracking is more challenging for the smallest joint stiffness. A gray-box model capturing the interactions of the two degrees of freedom is used in a learning scheme that is applied for the smallest stiffness level. The learning approach reduces interactions between the two degrees of freedom and demonstrates the control performance achievable with the system developed.

Numerical simulation of fault characteristics for refrigeration systems with liquid line receivers

Abstract Refrigeration systems can experience operating faults that negatively impact their performance, such as reducing their ability to maintain the temperature setpoint, increasing energy consumption, or reducing the life of the equipment. To better understand the impacts on the physical operating characteristics of a system, a numerical model can be used. However, fully physics-based models require significant computational effort and detailed system information that is typically unavailable; whereas data-driven models require a prohibitively large amount of test data. In this paper a gray-box model is described, using a semi-empirical approach from the literature. It combines into a complete cycle the models for each component: compressor, heat exchangers, pipelines, expansion valve, and liquid line receiver. A set of 35 test data from laboratory experiments on a commercial walk-in freezer was used for training and validating the model. The liquid line receiver was found to provide particular challenges for charge estimation, but the overall model performs quite well and is able to predict the effects of faults on most operating variables with reasonable accuracy.

An Online Grey-Box Model Based on Unscented Kalman Filter to Predict Temperature Profiles in Smart Buildings

Nearly 40% of primary energy consumption is related to the usage of energy in Buildings. Energy-related data such as indoor air temperature and power consumption of heating/cooling systems can be now collected due to the widespread diffusion of Internet-of-Things devices. Such energy data can be used (i) to train data-driven models than learn the thermal properties of buildings and (ii) to predict indoor temperature evolution. In this paper, we present a Grey-box model to estimate thermal dynamics in buildings based on Unscented Kalman Filter and thermal network representation. The proposed methodology has been applied in two different buildings with two different thermal network discretizations to test its accuracy in indoor air temperature prediction. Due to a lack of a real-world data sampled by Internet of Things (IoT) devices, a realistic data-set has been generated using the software Energy+, by referring to real industrial building models. Results on synthetic and realistic data show the accuracy of the proposed methodology in predicting indoor temperature trends up to the next 24 h with a maximum error lower than 2 °C, considering one year of data with different weather conditions.

PECLIDES Neuro: A Personalisable Clinical Decision Support System for Neurological Diseases.

Neurodegenerative diseases such as Alzheimer's and Parkinson's impact millions of people worldwide. Early diagnosis has proven to greatly increase the chances of slowing down the diseases' progression. Correct diagnosis often relies on the analysis of large amounts of patient data, and thus lends itself well to support from machine learning algorithms, which are able to learn from past diagnosis and see clearly through the complex interactions of a patient's symptoms and data. Unfortunately, many contemporary machine learning techniques fail to reveal details about how they reach their conclusions, a property considered fundamental when providing a diagnosis. Here we introduce our Personalisable Clinical Decision Support System (PECLIDES), an algorithmic process formulated to address this specific fault in diagnosis detection. PECLIDES provides a clear insight into the decision-making process leading to a diagnosis, making it a gray box model. Our algorithm enriches the fundamental work of Masheyekhi and Gras in data integration, personal medicine, usability, visualization, and interactivity.Our decision support system is an operation of translational medicine. It is based on random forests, is personalisable and allows a clear insight into the decision-making process. A well-structured rule set is created and every rule of the decision-making process can be observed by the user (physician). Furthermore, the user has an impact on the creation of the final rule set and the algorithm allows the comparison of different diseases as well as regional differences in the same disease. The algorithm is applicable to various decision problems. In this paper we will evaluate it on diagnosing neurological diseases and therefore refer to the algorithm as PECLIDES Neuro1.

Prediction of offshore wind farm power using a novel two-stage model combining kernel-based nonlinear extension of the Arps decline model with a multi-objective grey wolf optimizer

The development of the wind power market has led various countries to begin shifting the construction of wind farms from land to offshore. Accurately predicting the short-term wind power of an offshore wind farm is significant for preventive control and scheduling. This paper proposes a novel two-stage hybrid model to predict short-term wind power. First, complete ensemble empirical mode decomposition with adaptive noise is used to preprocess the raw data and make it smoother. Second, an improved grey-box model is utilized to predict the wind power, where a multi-objective grey wolf optimizer is used to optimize the kernel-based nonlinear extension of the Arps decline model to ensure both prediction stability and accuracy. The historical wind power data for offshore wind farms in Belgium are taken as a case study, and the results indicate that the proposed model has higher prediction accuracy and stability than those of six benchmark models. Three critical issues related to this work are also considered. The following conclusions are reached: (1) Utilizing a denoising method and multi-objective optimizer effectively improves the prediction accuracy and stability of the original model. Specifically, the multi-objective optimizer improves the prediction performance; (2) The proposed model performs best at predicting wind power in autumn; (3) Because of the great variation in dimensions for the wind power time series, the mean absolute percentage error and root mean squared percentage error are not suitable as error evaluation indicators.