William Holderbaum - Academia.edu (original) (raw)

Uploads

Papers by William Holderbaum

iScience

EAP gel memory mechanics were demonstrated via voltage potential measurements Probabilistic Moore... more EAP gel memory mechanics were demonstrated via voltage potential measurements Probabilistic Moore automata were constructed from EAP gel responses to stimulation Through tuning response encoding a computational reservoir was created The reservoir was shown as more memory efficient than general digital alternatives

Springer eBooks, 2023

The previous chapter introduced various concepts from statistics and probability relevant to fore... more The previous chapter introduced various concepts from statistics and probability relevant to forecasting. As many state-of-the-art approaches rely on machine learning, this chapter will introduce some fundamental definitions and concepts. It does not intend to provide an in-depth understanding but instead plans to overview the main concepts as they are relevant to this book. It provides an overview of practically relevant concepts when using software packages to fit and configure machine learning models. It does not introduce specific algorithms as those machine learning algorithms that are typically used in load forecasting are discussed in detail in Chap. 10. More in-depth overviews of the approaches can be found in the list of further reading in Appendix D.2.

Springer eBooks, 2023

Chapters 5 and 6 have shown how to define a time series forecast, how to prepare the data, and ho... more Chapters 5 and 6 have shown how to define a time series forecast, how to prepare the data, and how to generate inputs for the models. Chapters 9 to 11 will show several methods for forecasting the demand. However, although Chap. 7 provided us the tools for measuring the accuracy of a forecast, the following questions remain largely unanswered: How do we train and select a model which will consistently produce accurate forecasts? This chapter will investigate this question by looking at some of the most important aspects for creating a good forecast including proper utilisation of benchmarking, and how to use cross-validation to properly train your model. Underlying cross-validation is one of the most important aspects of a creating a good forecast, the so-called biasvariance trade-off principle, discussed in Sect. 8.1.2. This ensures that the model is not over (or under-) trained and allows the model to better generalise to new, unseen data. Next, in Sect. 8.2, methods for training the models are considered, including ways to select the best model from a selection of models. One important set of techniques covered in Sects. 8.2.4 and 8.2.5 is regularisation, which helps to reduce overfitting, but also how to find the appropriate hyperparameters within a family of models.

Springer eBooks, 2023

The previous sections have described in detail the steps required to develop a time series foreca... more The previous sections have described in detail the steps required to develop a time series forecast including: how to generate useful explanatory variables; how to train the model; how to avoid overfitting; and how to evaluate the accuracy of the model. What has not been investigated is the models themselves. This chapter will be the first of three chapters looking at a wide range of models and some of their properties. This chapter and the next will look at point forecast methods, and then in Chap. 11, probabilistic forecasts will be examined which provide models for handling highly uncertain data, something which is often required for low voltage feeders and substations (Chap. 2). Of the point forecasting chapters, this chapter looks at traditional statistical methods, whereas Chap. 10 will look at what are sometimes referred to as machine learning models. Each type of models has advantages and disadvantages, some of which have already been described in Sect. 5.3, but further criteria will be described in Sect. 12.2. In short, statistical models are typically more transparent and easier to interpret and understand. That makes them not only useful for investigating some of the core features of the data, but also makes them good benchmark candidates. The majority of the models presented in this chapter are easily implemented through packages in open source programming languages for scientific computing such as Python and R as well in popular proprietary software such as MATLAB. However they can be easily derived and trained from scratch (since they are often linear functions and hence can be easily trained using e.g. linear least squares, see Sect. 8.2), which may be preferable when you want to extend the models or make bespoke adjustments. This chapter starts by considering some simple models and then introduces progressively more complicated ones (in terms of more parameters and computational expense) starting with exponential smoothing (Sect. 9.2), multiple linear regression models (Sect. 9.3), ARIMA and SARIMA models (Sects. 9.4 and 9.5 respectively), and then finally generalised additive models (Sect. 9.6). Before diving into the models it is worth highlighting the context for these forecasts: short term load forecasts (STLF). A common way to categorise a load forecasts is in terms of the forecast horizon. Short term forecasts estimate the demand

Springer eBooks, 2023

This chapter demonstrates the practical implementation of short term (day-ahead) forecasts for th... more This chapter demonstrates the practical implementation of short term (day-ahead) forecasts for the application of residential low voltage networks. It is split into two main parts: An in-depth examination of a short term forecasting case study of residential low voltage networks (Sect. 14.2); and a example python code demonstrating how to implement some of the methods and techniques in practice (Sect. 14.3). The case studies serve to demonstrate how to: • identify the main challenges when implementing short term forecasts. • use the techniques from Chap. 6 to analyse the data, and identify important features. • use the analysis to choose several forecast models (from those presented in Chaps. 9 and 11). This includes both point and probabilistic models. • test, compare and evaluate the forecasts. The chapter begins by a short discussion of how to design a forecast trial which will frame the case study that follows later.

Accurate values of transmission line impedance parameters can enhance the efficiency and reliabil... more Accurate values of transmission line impedance parameters can enhance the efficiency and reliability of power system operation. The parameter values can be calculated using synchronized phasor measurements of voltage and current at the line ends. The objective of this paper is to investigate how the accuracy and uncertainty of the calculated parameter values depends on the level of noise and distortion in the measured signals during different line loading conditions. The investigation is conducted using a laboratory-based transmission line model and measurement system. The details of the test system and operation are presented in this paper.

Knowledge of real-time impedance parameters of overhead lines can significantly enhance power sys... more Knowledge of real-time impedance parameters of overhead lines can significantly enhance power system monitoring and control applications such as dynamic line rating and fault location. To ensure high accuracy of the determined parameters, methods for error detection and correction are required. In this paper the impact and correction of systematic errors in the voltage phasors are considered and corrected using an optimization procedure. The effectiveness of the proposed method was demonstrated in a case study involving a laboratory-based overhead line model. In contrast to a least squares-based estimator, a significant improvement in the impedance parameter accuracy was achieved, even when only phasor measurements from similar loading conditions were used.

European Journal of Control, 2004

International Journal of Electrical Power & Energy Systems, Mar 1, 2019

An Energy Storage System (ESS) is a potential solution to increase the energy efficiency of low v... more An Energy Storage System (ESS) is a potential solution to increase the energy efficiency of low voltage distribution networks whilst reinforcing the power system. In this article, energy management systems have been developed for the control of an ESS connected to a network of electrified Rubber Tyre Gantry (RTG) cranes. Considering the highly volatile crane demand behaviour and uncertainty in the RTG crane demand prediction as a nonlinear optimisation problem, this paper presents and verifies an optimal energy control strategy based on a Stochastic Model Predictive Control (SMPC) algorithm. The SMPC controller aims to improve the reliability and economic performance of a network of RTG cranes, under a given ESS and network specification. A specific case, using different ESS locations, is presented and the results of the proposed SMPC and MPC control models are compared to a set-point controller using data collected from an instrumented electrified RTG cranes at the Port of Felixstowe, UK. The results indicate that the SMPC controller successfully reduce electrical energy costs, the peak demand and outperforms each of the presented control techniques.

iScience

EAP gel memory mechanics were demonstrated via voltage potential measurements Probabilistic Moore... more EAP gel memory mechanics were demonstrated via voltage potential measurements Probabilistic Moore automata were constructed from EAP gel responses to stimulation Through tuning response encoding a computational reservoir was created The reservoir was shown as more memory efficient than general digital alternatives

Springer eBooks, 2023

The previous chapter introduced various concepts from statistics and probability relevant to fore... more The previous chapter introduced various concepts from statistics and probability relevant to forecasting. As many state-of-the-art approaches rely on machine learning, this chapter will introduce some fundamental definitions and concepts. It does not intend to provide an in-depth understanding but instead plans to overview the main concepts as they are relevant to this book. It provides an overview of practically relevant concepts when using software packages to fit and configure machine learning models. It does not introduce specific algorithms as those machine learning algorithms that are typically used in load forecasting are discussed in detail in Chap. 10. More in-depth overviews of the approaches can be found in the list of further reading in Appendix D.2.

Springer eBooks, 2023

Chapters 5 and 6 have shown how to define a time series forecast, how to prepare the data, and ho... more Chapters 5 and 6 have shown how to define a time series forecast, how to prepare the data, and how to generate inputs for the models. Chapters 9 to 11 will show several methods for forecasting the demand. However, although Chap. 7 provided us the tools for measuring the accuracy of a forecast, the following questions remain largely unanswered: How do we train and select a model which will consistently produce accurate forecasts? This chapter will investigate this question by looking at some of the most important aspects for creating a good forecast including proper utilisation of benchmarking, and how to use cross-validation to properly train your model. Underlying cross-validation is one of the most important aspects of a creating a good forecast, the so-called biasvariance trade-off principle, discussed in Sect. 8.1.2. This ensures that the model is not over (or under-) trained and allows the model to better generalise to new, unseen data. Next, in Sect. 8.2, methods for training the models are considered, including ways to select the best model from a selection of models. One important set of techniques covered in Sects. 8.2.4 and 8.2.5 is regularisation, which helps to reduce overfitting, but also how to find the appropriate hyperparameters within a family of models.

Springer eBooks, 2023

The previous sections have described in detail the steps required to develop a time series foreca... more The previous sections have described in detail the steps required to develop a time series forecast including: how to generate useful explanatory variables; how to train the model; how to avoid overfitting; and how to evaluate the accuracy of the model. What has not been investigated is the models themselves. This chapter will be the first of three chapters looking at a wide range of models and some of their properties. This chapter and the next will look at point forecast methods, and then in Chap. 11, probabilistic forecasts will be examined which provide models for handling highly uncertain data, something which is often required for low voltage feeders and substations (Chap. 2). Of the point forecasting chapters, this chapter looks at traditional statistical methods, whereas Chap. 10 will look at what are sometimes referred to as machine learning models. Each type of models has advantages and disadvantages, some of which have already been described in Sect. 5.3, but further criteria will be described in Sect. 12.2. In short, statistical models are typically more transparent and easier to interpret and understand. That makes them not only useful for investigating some of the core features of the data, but also makes them good benchmark candidates. The majority of the models presented in this chapter are easily implemented through packages in open source programming languages for scientific computing such as Python and R as well in popular proprietary software such as MATLAB. However they can be easily derived and trained from scratch (since they are often linear functions and hence can be easily trained using e.g. linear least squares, see Sect. 8.2), which may be preferable when you want to extend the models or make bespoke adjustments. This chapter starts by considering some simple models and then introduces progressively more complicated ones (in terms of more parameters and computational expense) starting with exponential smoothing (Sect. 9.2), multiple linear regression models (Sect. 9.3), ARIMA and SARIMA models (Sects. 9.4 and 9.5 respectively), and then finally generalised additive models (Sect. 9.6). Before diving into the models it is worth highlighting the context for these forecasts: short term load forecasts (STLF). A common way to categorise a load forecasts is in terms of the forecast horizon. Short term forecasts estimate the demand

Springer eBooks, 2023

This chapter demonstrates the practical implementation of short term (day-ahead) forecasts for th... more This chapter demonstrates the practical implementation of short term (day-ahead) forecasts for the application of residential low voltage networks. It is split into two main parts: An in-depth examination of a short term forecasting case study of residential low voltage networks (Sect. 14.2); and a example python code demonstrating how to implement some of the methods and techniques in practice (Sect. 14.3). The case studies serve to demonstrate how to: • identify the main challenges when implementing short term forecasts. • use the techniques from Chap. 6 to analyse the data, and identify important features. • use the analysis to choose several forecast models (from those presented in Chaps. 9 and 11). This includes both point and probabilistic models. • test, compare and evaluate the forecasts. The chapter begins by a short discussion of how to design a forecast trial which will frame the case study that follows later.

Accurate values of transmission line impedance parameters can enhance the efficiency and reliabil... more Accurate values of transmission line impedance parameters can enhance the efficiency and reliability of power system operation. The parameter values can be calculated using synchronized phasor measurements of voltage and current at the line ends. The objective of this paper is to investigate how the accuracy and uncertainty of the calculated parameter values depends on the level of noise and distortion in the measured signals during different line loading conditions. The investigation is conducted using a laboratory-based transmission line model and measurement system. The details of the test system and operation are presented in this paper.

Knowledge of real-time impedance parameters of overhead lines can significantly enhance power sys... more Knowledge of real-time impedance parameters of overhead lines can significantly enhance power system monitoring and control applications such as dynamic line rating and fault location. To ensure high accuracy of the determined parameters, methods for error detection and correction are required. In this paper the impact and correction of systematic errors in the voltage phasors are considered and corrected using an optimization procedure. The effectiveness of the proposed method was demonstrated in a case study involving a laboratory-based overhead line model. In contrast to a least squares-based estimator, a significant improvement in the impedance parameter accuracy was achieved, even when only phasor measurements from similar loading conditions were used.

European Journal of Control, 2004

International Journal of Electrical Power & Energy Systems, Mar 1, 2019

An Energy Storage System (ESS) is a potential solution to increase the energy efficiency of low v... more An Energy Storage System (ESS) is a potential solution to increase the energy efficiency of low voltage distribution networks whilst reinforcing the power system. In this article, energy management systems have been developed for the control of an ESS connected to a network of electrified Rubber Tyre Gantry (RTG) cranes. Considering the highly volatile crane demand behaviour and uncertainty in the RTG crane demand prediction as a nonlinear optimisation problem, this paper presents and verifies an optimal energy control strategy based on a Stochastic Model Predictive Control (SMPC) algorithm. The SMPC controller aims to improve the reliability and economic performance of a network of RTG cranes, under a given ESS and network specification. A specific case, using different ESS locations, is presented and the results of the proposed SMPC and MPC control models are compared to a set-point controller using data collected from an instrumented electrified RTG cranes at the Port of Felixstowe, UK. The results indicate that the SMPC controller successfully reduce electrical energy costs, the peak demand and outperforms each of the presented control techniques.