E Elena Mocanu

A topological insight into restricted Boltzmann machines

Restricted Boltzmann Machines (RBMs) and models derived from them have been successfully used as basic building blocks in deep artificial neural networks for automatic features extraction, unsupervised weights initialization, but also as density estimators. Thus, their generative and discriminative capabilities, but also their computational time are instrumental to a wide range of applications. Our main contribution is to look at RBMs from a topological perspective, bringing insights from network science. Firstly, here we show that RBMs and Gaussian RBMs (GRBMs) are bipartite graphs which naturally have a small-world topology. Secondly, we demonstrate both on synthetic and real-world datasets that by constraining RBMs and GRBMs to a scale-free topology (while still considering local neighborhoods and data distribution), we reduce the number of weights that need to be computed by a few orders of magnitude, at virtually no loss in generative performance. Thirdly, we show that, for a fixed number of weights, our proposed sparse models (which by design have a higher number of hidden neurons) achieve better generative capabilities than standard fully connected RBMs and GRBMs (which by design have a smaller number of hidden neurons), at no additional computational costs.

Comparison of machine learning methods for estimating energy consumption in buildings

The increasing number of decentralized renewable energy sources together with the grow in overall electricity consumption introduce many new challenges related to dimensioning of grid assets and supply-demand balancing. Approximately 40% of the total energy consumption is used to cover the needs of commercial and office buildings. To improve the design of the energy infrastructure and the efficient deployment of resources, new paradigms have to be thought up. Such new paradigms need automated methods to dynamically predict the energy consumption in buildings. At the same time these methods should be easily expandable to higher levels of aggregation such as neighborhoods and the power grid. Predicting energy consumption for a building is complex due to many influencing factors, such as weather conditions, performance and settings of heating and cooling systems, and the number of people present. In this paper, we investigate a newly developed stochastic model for time series prediction of energy consumption, namely the Conditional Restricted Boltzmann Machine (CRBM), and evaluate its performance in the context of building automation systems. The assessment is made on a real dataset consisting of 7 weeks of hourly resolution electricity consumption collected from a Dutch office building. The results showed that for the energy prediction problem solved here, CRBM outperforms Artificial Neural Networks (ANNs), and Hidden Markov Models (HMMs).

Comfort-constrained demand flexibility management for building aggregations using a decentralized approach

In the smart grid and smart city context, the energy end-user plays an active role in the operation of the power system. The rapid penetration of Renewable Energy Sources (RES) and Distributed Energy Resources (DER) requires a higher degree of flexibility on the demand side. As commercial and Industrial buildings (C&I) buildings represent a substantial aggregation of loads, the intertwined operation of the electric distribution network and the built environment is to large extent responsible for achieving energy efficiency and sustain-ability targets. However, the primary purpose of buildings is not grid support but rather ensuring the comfort and safety of its occupants. Therefore, the comfort level needs to be included as a constraint when assessing the flexibility potential of the built environment. This paper proposes a decentralized method for flexibility allocation among a set of buildings. The method uses concepts from non-cooperative game theory. Finally, two case of study are used to evaluate the performance of the decentralized algorithm, and compare it against a centralized option. It is shown that flexibility requests from the grid operator can be met without deteriorating the comfort levels.

Optimizing the energy exchange between the Smart Grid and Building Systems

Various Smart Grid (SG) technologies and concepts are currently under investigation, driven by the goals of energy transition policies towards future sustainable, reliable and affordable electricity supply systems. This paper presents an approach for modeling the interaction between the Smart Grid and Building Energy Management Systems (SG-BEMS), using Multi Agent Systems control. The interaction consists of three layers: the smart building, the neighborhood, and the distribution grid. It enables the continuous bidirectional flow of energy and information between SG and BEMS. The proposed framework combines optimization techniques inspired by dynamic game theory and stochastic optimization algorithms. The goal of the optimization is to increase the overall performance, while keeping a good level of comfort for people in the built environment.

Inexpensive user tracking using Boltzmann machines

Inexpensive user tracking is an important problem in various application domains such as healthcare, human-computer interaction, energy savings, safety, robotics, security and so on. Yet, it cannot be easily solved due to its probabilistic nature, high level of abstraction and uncertainties, on the one side, and to the limitations of our current technologies and learning algorithms, on the other side. In this paper, we tackle this problem by using the Multi-integrated Sensor Technology, which comes at a low price. At the same time, we are aiming to address the lightweight learning requirements by investigating Factored Conditional Restricted Boltzmann Machines (FCRBMs), a form of Deep Learning, that has proven to be an efficient and effective machine learning framework. However, due to their construction properties, the conventional FCRBMs are only capable of performing predictions but are not capable of making classification. Herein, we are proposing extended FCRBMs (eFCRBMs), which incorporate a novel classification scheme, to solve this problem. Experiments performed on both artificially generated as well as real-world data demonstrate the effectiveness and efficiency of the proposed technique. We show that eFCRBMs outperform popular approaches including Support Vector Machines, Naive Bayes, AdaBoost, and Gaussian Mixture Models.

Big IoT data mining for real-time energy disaggregation in buildings

In the smart grid context, the identification and prediction of building energy flexibility is a challenging open question, thus paving the way for new optimized behaviors from the demand side. At the same time, the latest smart meters developments allow us to monitor in real-time the power consumption level of the home appliances, aiming at a very accurate energy disaggregation. However, due to practical constraints is infeasible in the near future to attach smart meter devices on all home appliances, which is the problem addressed herein. We propose a hybrid approach, which combines sparse smart meters with machine learning methods. Using a subset of buildings equipped with subset of smart meters we can create a database on which we train two deep learning models, i.e. Factored Four-Way Conditional Restricted Boltzmann Machines (FFW-CRBMs) and Disjunctive FFW-CRBM. We show how our method may be used to accurately predict and identify the energy flexibility of buildings unequipped with smart meters, starting from their aggregated energy values. The proposed approach was validated on a real database, namely the Reference Energy Disaggregation Dataset. The results show that for the flexibility prediction problem solved here, Disjunctive FFW-CRBM outperforms the FFW-CRBMs approach, where for classification task their capabilities are comparable.

Energy disaggregation for real-time building flexibility detection

Energy is a limited resource which has to be managed wisely, taking into account both supply-demand matching and capacity constraints in the distribution grid. One aspect of the smart energy management at the building level is given by the problem of real-time detection of flexible demand available. In this paper we propose the use of energy disaggregation techniques to perform this task. Firstly, we investigate the use of existing classification methods to perform energy disaggregation. A comparison is performed between four classifiers, namely Naive Bayes, k-Nearest Neighbors, Support Vector Machine and AdaBoost. Secondly, we propose the use of Restricted Boltzmann Machine to automatically perform feature extraction. The extracted features are then used as inputs to the four classifiers and consequently shown to improve their accuracy. The efficiency of our approach is demonstrated on a real database consisting of detailed appliance-level measurements with high temporal resolution, which has been used for energy disaggregation in previous studies, namely the REDD. The results show robustness and good generalization capabilities to newly presented buildings with at least 96% accuracy.

The potential and possible effects of power grid support activities on buildings: An analysis of experimental results for ventilation system

This paper reports on the potential and possible effects of using building services installations (notably ventilation systems) to support power grids. This is significant taken that the shift towards smart grids comes with adoption of demand side integration and the concept of active controllable loads. However, it is recommended that demand side resource will be used for grid support activities only if non-disruption in terms of indoor comfort and their responsiveness can be guaranteed. Relevant studies mainly report grid perspective in event of using demand side resources to support the power grid. The result is that little emphasis is given to indoor comfort, building behavior and the exact details of achieving controllability at building level in such events. Using experimental data from an office building in the Netherlands this paper reports on indoor comfort and building behavior in the event of committing installed ventilation systems to provide power grid support services. Possibilities for attaining controllability and responsiveness for the components in such systems are also presented. The study is case specific and contributes to the development of possible operational guidelines for building ventilation systems in event of using them for grid support activities.

Demand forecasting at low aggregation levels using Factored Conditional Restricted Boltzmann Machine

The electrical demand forecasting problem can be regarded as a non-linear time series prediction problem depending on many complex factors since it is required at various aggregation levels and at high resolution. To solve this challenging problem, various time series and machine learning approaches has been proposed in the literature. As an evolution of neural network-based prediction methods, deep learning techniques are expected to increase the prediction accuracy by being stochastic and allowing bi-directional connections between neurons. In this paper, we investigate a newly developed deep learning model for time series prediction, namely Factored Conditional Restricted Boltzmann Machine (FCRBM), and extend it for demand forecasting. The assessment is made on the EcoGrid EU dataset, consisting of aggregated electric power consumption, price and meteorological data collected from 1900 customers. The households are equipped with local generation and smart appliances capable of responding to real-time pricing signals. The results show that for the energy prediction problem solved here, FCRBM outperforms the benchmark machine learning approach, i.e. Support Vector Machine.

Medium Voltage DC Power Systems on Ships: An Offline Parameter Estimation for Tuning the Controllers’ Linearizing Function

Future shipboard power systems using Medium Voltage Direct (MVDC) technology will be based on a widespread use of power converters for interfacing generating systems and loads with the main DC bus. Such a heavy exploitation makes the voltage control challenging in the presence of tightly controlled converters. By modeling the latter as constant power loads (CPLs), one possibility to ensure the bus voltage stability is offered by the linearizing via state feedback technique, whose aim is to regulate the generating DC-DC power converters to compensate for the destabilizing effect of the CPLs. Although this method has been shown to be effective when system parameters are perfectly known, only a partial linearization can be ensured in case of parameter mismatch, thus, jeopardizing the system stability. In order to improve the linearization, therefore, guaranteeing the voltage stability, an estimation method is proposed in this paper. To this aim, offline tests are performed to provide the input data for the estimation of model parameters. Such estimated values are subsequently used for correctly tuning the linearizing function of the DC-DC converters. Simulation results for bus voltage transients show that in this way converters become sources of stabilizing power.