A Parallel Evolutionary Computing-Embodied Artificial Neural Network Applied to Non-Intrusive Load Monitoring for Demand-Side Management in a Smart Home: Towards Deep Learning

Non-intrusive load monitoring (NILM) is a cost-effective approach that electrical appliances are identified from aggregated whole-field electrical signals, according to their extracted electrical characteristics, with no need to intrusively deploy smart power meters (power plugs) installed for individual monitored electrical appliances in a practical field of interest. This work addresses NILM by a parallel Genetic Algorithm (GA)-embodied Artificial Neural Network (ANN) for Demand-Side Management (DSM) in a smart home. An ANN’s performance in terms of classification accuracy depends on its training algorithm. Additionally, training an ANN/deep NN learning from massive training samples is extremely computationally intensive. Therefore, in this work, a parallel GA has been conducted and used to integrate meta-heuristics (evolutionary computing) with an ANN (neurocomputing) considering its evolution in a parallel execution relating to load disaggregation in a Home Energy Management System (HEMS) deployed in a real residential field. The parallel GA that involves iterations to excessively cost its execution time for evolving an ANN learning model from massive training samples to NILM in the HEMS and works in a divide-and-conquer manner that can exploit massively parallel computing for evolving an ANN and, thus, reduce execution time drastically. This work confirms the feasibility and effectiveness of the parallel GA-embodied ANN applied to NILM in the HEMS for DSM.

A New Parallel GA-Based Method for Constraint Satisfaction Problems

Despite some success of Genetic Algorithms (GAs) when tackling Constraint Satisfaction Problems (CSPs), they generally suffer from poor crossover operators. In order to overcome this limitation in practice, we propose a novel crossover specifically designed for solving CSPs including Temporal CSPs (TCSPs). Together with a variable ordering heuristic and an integration into a parallel architecture, this proposed crossover enables the solving of large and hard problem instances as demonstrated by the experimental tests conducted on randomly generated CSPs and TCSPs based on the model RB. We will indeed demonstrate, through these tests, that our proposed method is superior to the known GA-based techniques for CSPs. In addition, we will show that we are able to compete with the efficient MAC-based Abscon 109 solver for random problem instances as well as those instances taken from Lecoutre’s CSP library. Finally, we conducted additional tests on very large consistent and over constrained CSPs and TCSPs instances in order to show the ability of our method to deal with constraint problems in real time. This corresponds to solving the CSP or the TCSP by giving a solution with a quality (number of solved constraints) depending on the time allocated for computation.

Traffic Flow Prediction Model for Large-Scale Road Network Based on Cloud Computing

To increase the efficiency and precision of large-scale road network traffic flow prediction, a genetic algorithm-support vector machine (GA-SVM) model based on cloud computing is proposed in this paper, which is based on the analysis of the characteristics and defects of genetic algorithm and support vector machine. In cloud computing environment, firstly, SVM parameters are optimized by the parallel genetic algorithm, and then this optimized parallel SVM model is used to predict traffic flow. On the basis of the traffic flow data of Haizhu District in Guangzhou City, the proposed model was verified and compared with the serial GA-SVM model and parallel GA-SVM model based on MPI (message passing interface). The results demonstrate that the parallel GA-SVM model based on cloud computing has higher prediction accuracy, shorter running time, and higher speedup.