This chapter contains sections titled: Historical Review Supervised Multilayer Networks Unsupervised Neural Networks: Kohonen Network Unsupervised Networks: Adaptive Resonance Theory Network Model Validation Summary References Recommended Exercises

Nonlinear System Identification

Wireless body area networks (WBANs) form a new and interesting area in the world of remote health monitoring. An important concern in such networks is the communication between the sensors. This communication needs to be energy efficient and highly reliable while keeping delays low. Mobility also has to be supported as the nodes are positioned on different parts of the body that move with regard to each other. In this paper, we present a new cross-layer communication protocol for WBANs: CICADA or Cascading Information retrieval by Controlling Access with Distributed slot Assignment. The protocol sets up a network tree in a distributed manner. This tree structure is subsequently used to guarantee collision free access to the medium and to route data towards the sink. The paper analyzes CICADA and shows simulation results. The protocol offers low delay and good resilience to mobility. The energy usage is low as the nodes can sleep in slots where they are not transmitting or receiving.

A low-delay protocol for multishop wireless body area networks

Adaptive receding horizon control for constrained MIMO systems

Rolling element bearings are among the most widely used and also vulnerable components in rotating machinery equipment. Recently, prognostics and health management of rolling element bearings is more and more attractive both in academics and industry. However, many studies have been focusing on the prognostic aspect of bearing prognostics and health management and few efforts have been performed in relation to the optimal degradation feature selection issue. For more effective and efficient remaining useful life predictions, three goodness metrics of correlation, monotonicity and robustness are defined and combined for automatically more relevant degradation feature selection in this paper. Effectiveness of the proposed method is verified by rolling element bearing degradation experiments. Copyright © 2015 John Wiley & Sons, Ltd.

Degradation Feature Selection for Remaining Useful Life Prediction of Rolling Element Bearings

The field of photovoltaics gives the opportunity to make our buildings "smart'' and our portable devices "independent", provided effective energy sources can be developed for use in ambient indoor ...

/pdf/dye-sensitized-solar-cells-under-ambient-light-powering-1yxiu7v3jw.pdf

Dye-sensitized solar cells under ambient light powering machine learning : towards autonomous smart sensors for the internet of things

This paper presents a data fusion system, which models and handles uncertainties in buildings, in order to design a smart heating system. The data fusion model provides a temperature trend that will be inserted in smart and predictive heating systems in order to reduce energy consumption in the building. The multilevel data fusion model is based on the belief functions theory. Different simulations were carried out and give very satisfactory results, in spite of the complexity of the data.

http://ieeexplore.ieee.org/iel7/7502235/7507016/07507074.pdf

Building heating management based on the belief functions theory

This paper proposes the design and a comparative study of two non linear Multiple Inputs Single Output (MISO) models. The first, titled Volterra model, is built using Volterra series and the second, named RKHS model, uses the Statistical Learning Theory (SLT) which operates on Reproducing Kernel Hilbert Space (RKHS). The complexity of both models is pointed out in SISO models as well as in MISO ones. The performances of both models are evaluated first by using Monte Carlo numerical simulations and second by handling a chemical process known as the Continuous Stirred Tank Reactor CSTR. In both validation operation the results were successful.

A comparative study of non linear MISO Process modelling techniques: Application to a chemical reactor

When prior information on plant behaviour are not available, recursive parameter bounding algorithms often use loose initial bounds which may lead to a large final membership sets. Conversely if bounds on some physically meaningful parameters of the dynamical system are available they help to compute prior bounds that can be used as an initial feasible set which accelerate the algorithm convergence and contribute largely to yield a narrower final membership set. In this paper we develop a method which compute the initial feasible set when bounds on the damping factor, the natural pulsation and the steady state gain are available. The computed feasible set is used as initial bounds on polytope updating algorithm. The volume evolution of the feasible set as well as the activity of constraints defining the initial bounds, during updating operation, are compared to the case of loose initial bounds.

Hassani Messaoud

Papers

Building heating management based on the belief functions theory

A comparative study of non linear MISO Process modelling techniques: Application to a chemical reactor

An algorithm for computing parameter bounds using prior information on physical parameter bounds