Genetic algorithms for local controller network construction

Local Controller Networks (LCNs) provide nonlinear control by interpolating between a set of locally valid, subcontrollers covering the operating range of the plant. Constructing such networks typically requires knowledge of valid local models. This paper describes a new genetic learning approach to the construction of LCNs directly from the dynamic equations of the plant, or from modelling data. The advantage is that a priori knowledge about valid local models is not needed. In addition to allowing simultaneous optimisation of both the controller and validation function parameters, the approach aids transparency by ensuring that each local controller acts independently of the rest at its operating point. It thus is valuable for simultaneous design of the LCNs and identification of the operating regimes of an unknown plant. Application results from a highly nonlinear pH neutralisation process and its associated neural network representation are utilised to illustrate these issues.

Comparison of two construction algorithms for local model networks

The divide-and-conquer approach of local model (LM) networks is a common engineering approach to the identification of a complex nonlinear dynamical system. The global representation is obtained from the weighted sum of locally valid, simpler sub-models defined over small regions of the operating space. Constructing such networks requires the determination of appropriate partitioning and the parameters of the LMs. This paper focuses on the structural aspect of LM networks. It compares the computational requirements and performances of the Johansen and Foss (J&F) and LOLIMOT tree-construction algorithms. Several useful and important modifications to each algorithm are proposed. The modelling performances are evaluated using real data from a pilot plant of a pH neutralization process. Results show that while J&F achieves a more accurate nonlinear representation of the pH process, LOLIMOT requires significantly less computational effort.

Monoamines and decision making during contests in the hermit crab Pagurus bernhardus

During fights animals are expected to make a series of strategic decisions that involve interactions between information about the contest and the individual's nervous system that produce a change in behaviour. Biogenic monoamines such as serotonin ('5-HT') and dopamine are thought to prime decision-making centres for appropriate responses during aggressive interactions in crustaceans, and circulating levels vary both between individuals and during agonistic encounters. Aminergenic systems operate in diverse animal taxa and in this study we assayed circulating levels of S-HT and dopamine following shell fights in the common European hermit crab, Pagurus bernhardus. The two roles in these fights, attacker and defender, perform different activities but, in both, S-HT increased and dopamine declined in response to engaging in a fight. In defenders but not attackers, giving up was correlated with low 5-HT and dopamine. In attackers, motivation to initiate a fight was positively correlated with dopamine levels. Circulating monoamines are therefore involved in decision making during these aggressive encounters. (c) 2007 The Association for the Study of Animal Behaviour Published by Elsevier Ltd. All rights reserved.

Self-assessment by males during energetically costly contests over precopula females in amphipods

In animal contests, individuals can either engage in mutual assessment of both their own and their opponent's resource-holding potential (RHP) and adjust their behaviour according to estimated differences, or instead persist in accordance with thresholds determined by assessment of just their own RHP. We examined the predictions of alternative mutual assessment and self-assessment models for decision rules in contest resolution during struggles between males over females in precopula in the amphipod Gammarus pulex. Contest duration was positively related to the weight of the loser but not the weight of the winner. Our results support the hypothesis that males rely on information about their own RHP in determining contest behaviour and do not use information about their opponent. Fighting was energetically costly, and energy reserves were depleted during contests. Contest duration was associated with the physiological state of the loser (but not the winner) at the end of the contest, and to a lesser extent his size, further supporting self-assessment. (c) 2006 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.