Katya Rodríguez-Vázquez

Identifying the structure of nonlinear dynamic systems using multiobjective genetic programming

A method for identifying the structure of nonlinear polynomial dynamic models is presented. This approach uses an evolutionary algorithm, genetic programming, in a multiobjective fashion to generate global models which describe the dynamic behavior of the nonlinear system under investigation. The validation stage of system identification is simultaneously evaluated using the multiobjective tool, in order to direct the identification process to a set of global models of the system.

Application of system identification techniques to aircraft gas turbine engines

Abstract A variety of system identification techniques are applied to the modelling of aircraft gas turbine dynamics. The motivation behind the study is to improve the efficiency and cost-effectiveness of system identification techniques currently used in the industry. Three system identification approaches are outlined in this paper. They are based upon: multisine testing and frequency-domain identification, time-varying models estimated using extended least squares with optimal smoothing, and multiobjective genetic programming to select model structure.

Nonlinear identification of aircraft gas-turbine dynamics

Abstract Identification results for the shaft-speed dynamics of an aircraft gas turbine, under normal operation, are presented. As it has been found that the dynamics vary with the operating point, nonlinear models are employed. Two different approaches are considered: NARX models, and neural network models, namely multilayer perceptrons, radial basis function networks and B-spline networks. A special attention is given to genetic programming, in a multiobjective fashion, to determine the structure of NARMAX and B-spline models.

System identification strategies applied to aircraft gas turbine engines

Abstract A variety of system identification techniques are applied to the derivation of models of aircraft gas turbine dynamics. The motivation behind the study is to improve the efficiency and cost-effectiveness of system identification techniques currently used in the industry. Four system identification approaches are outlined in this paper. They are based upon: identification using ambient noise only data, multisine testing and frequency-domain identification, time-varying models estimated using extended least squares with optimal smoothing, and multiobjective genetic programming to select model structure.

Evolution of mathematical models of chaotic systems based on multiobjective genetic programming

This work is concerned with the identification of models for nonlinear dynamical systems using multiobjective evolutionary algorithms. Systems modelling involves the processes of structure selection, parameter estimation, model performance and model validation and involves a complex solution space. Evolutionary Algorithms (EAs) are search and optimisation tools founded on the principles of natural evolution and genetics, which are suitable for a wide range of application areas. Due to the versatility of these tools and motivated by the versatility of genetic programming (GP), this evolutionary paradigm is proposed for this modelling problem. GP is then combined with a multiobjective function definition scheme. Multiobjective genetic programming (MOGP) is applied to multiple, conflicting objectives and yields a set of candidate parsimonious and valid models, which reproduce the original system behaviour. The MOGP approach is then demonstrated as being applicable for system modelling with chaotic dynamics. The circuit introduced by Chua, being one of the most popular benchmarks for studying nonlinear oscillations, and the Duffing–Holmes oscillator are the systems to test the evolutionary-based modelling approach introduced in this paper.

Increments and duplication events of enzymes and transcription factors influence metabolic and regulatory diversity in prokaryotes.

In this work, the content of enzymes and DNA-binding transcription factors (TFs) in 794 non-redundant prokaryotic genomes was evaluated. The identification of enzymes was based on annotations deposited in the KEGG database as well as in databases of functional domains (COG and PFAM) and structural domains (Superfamily). For identifications of the TFs, hidden Markov profiles were constructed based on well-known transcriptional regulatory families. From these analyses, we obtained diverse and interesting results, such as the negative rate of incremental changes in the number of detected enzymes with respect to the genome size. On the contrary, for TFs the rate incremented as the complexity of genome increased. This inverse related performance shapes the diversity of metabolic and regulatory networks and impacts the availability of enzymes and TFs. Furthermore, the intersection of the derivatives between enzymes and TFs was identified at 9,659 genes, after this point, the regulatory complexity grows faster than metabolic complexity. In addition, TFs have a low number of duplications, in contrast to the apparent high number of duplications associated with enzymes. Despite the greater number of duplicated enzymes versus TFs, the increment by which duplicates appear is higher in TFs. A lower proportion of enzymes among archaeal genomes (22%) than in the bacterial ones (27%) was also found. This low proportion might be compensated by the interconnection between the metabolic pathways in Archaea. A similar proportion was also found for the archaeal TFs, for which the formation of regulatory complexes has been proposed. Finally, an enrichment of multifunctional enzymes in Bacteria, as a mechanism of ecological adaptation, was detected.

Autonomous Demand-Side Management system based on Monte Carlo Tree Search

Smart Grid (SG) technologies are becoming increasingly dynamic, motivating the use of computational intelligence to support the SG by predicting and intelligently responding to certain requests (e.g, reducing electricity costs given fluctuating prices). The presented work intends to do precisely this, to make intelligent decisions to switch on electric devices at times when the electricity price (prices that change over time) is the lowest while at the same time attempting to balance energy usage by avoiding turning on multiple devices at the same time, whenever possible. To this end, we use Monte Carlo Tree Search (MCTS), a real-time decision algorithm. MCTS takes into consideration what might happen in the future by approximating what other entities/agents (electric devices) might do via Monte Carlo simulations. We propose two variants of this method: (a) maxn MCTS approach where the competition for resources (e.g, lowest electricity price) happens in one single decision tree and where all the devices are considered, and (b) two-agent MCTS approach, where the competition for resources is distributed among various decision trees. To validate our results, we used two scenarios, a rather simple one where there are no constraints associated to the problem, and another more complex, and realistic scenario with equality and inequality constraints associated to the problem. The results achieved by this real-time decision tree algorithm are very promising, specially those achieved by the maxn MCTS approach.

An Evolutionary Approach to Non-Linear Polynomial System Identification

Abstract This work presents a genetic programming approach to the identification of polynomial models for non-linear systems. The genetic approach optimises the Akaike Information Criterion (AIC) in order to find the model structure and estimate the parameters. This includes a measures of the number of terms in the model which can be of varying degree and lag.

Genetic programming for dynamic chaotic systems modelling

This work presents an investigation into the use of genetic programming (GP) applied to chaotic systems modelling. A difference equation model representation was proposed for being the basis of the hierarchical tree encoding in GP. Based upon the NARMA difference equation model and formulating the identification as a multiobjective optimisation problem, Chuas circuit was studied. The formulation of the GP fitness function, defined as a multiobjective function, generated a set of nondominated chaotic models. This approach considered criteria related to the complexity, performance and also statistical validation of the models in the fitness evaluation. The final set of non-dominated model solutions were able to capture the dynamic characteristics of the system and reproduce the chaotic motion of the double scroll attractor.

Parallel particle swarm optimization applied to the protein folding problem

This article presents the implementation of a bio-inspired algorithm, which is the algorithm of particle swarm optimization (PSO) in with the objective of minimizing the function of conformational energy ECEPP/3 for the protein folding problem (PFP) for real conformations considering structural restrictions. In this case, using a representation of torsion angles of the skeleton and the side chains, applying the sequence of amino acid of the peptide leu-enkephalin for the prediction of 3D structure of minimum energy. The quality of the results is compared with other techniques reported in literature. Subsequently, the PSO is used to predict the structure of unknown proteins.