(12d) A Novel, Biomimetic Approach to Self-Organizing, Optimal Control Structure Design

While there are numerous works on controller design, research in the area of optimal control structure design is nascent. Traditionally, heuristics and process knowledge have been employed for control structure design. More recently, methodical approaches have been proposed (Halvorsen et al., 2003) based on economics. Optimal selection of the primary controlled variables (CVs) by considering plant economics as well as controllability in presence of closed loop dynamics and time delay has been recently proposed by some of the co-authors of this work (Jones et al., 2014a, 2014b). Due to large number of candidate variables that lead to significant number of combinatorial possibilities even for small plants, online selection of primary CVs is computationally prohibitive. Additionally, as process dynamics, operational constraints and operational objectives change, it is expected that optimal controlled variables would change. This necessitates online selection of CVs leading to a self-organizing optimal control structure. However, the large scale combinatorial problem that is undertaken for CV selection is computationally prohibitive for online application. Therefore, we propose a novel biomimetic approach that exploits the structural connectivity of the plant, similar to the self-organization of the cortical/sub-cortical areas in human brain, to decompose the large-scale problem into smaller subsets called ‘islands’. This approach makes parallel solution of the controlled variable selection problem feasible thus achieving solution within the target execution time. Secondly, within each island, a metaheuristic multiagent optimizer, that mimics the human nervous system, is deployed to obtain optimal CVs subject to the time constraints for online applications. The biomimetic approach addresses the problem of sub-optimality of controlled variables due to system changes and render the computational time tractable for online deployment.

Firstly, for determining the change in the connectivity among plant sections, we draw from the neuroscience literature where the connectivity among cortical/sub-cortical areas are determined using the fMRI responses (Friston et al., 2003). At a cortical level, the neuronal populations can be modelled as states that dynamically evoke brain responses as a function of inputs. Thus, by modeling islands of sub processes in a chemical plant as cortical/sub-cortical areas, the effective connectivity of these islands of sub processes can be extracted. In this approach, a bilinear model of the process is developed where the model parameters represent intrinsic coupling among the states, describe the influence of extrinsic inputs on the states and as well as capture the effect of inputs on coupling. These parameters are identified using a Bayesian framework. These connectivity matrices can then be employed as a basis for controlled variable selection.

Secondly, for each identified island, controlled variables are selected via multi-agent optimization techniques that employ various probabilistic and mathematical programming algorithms. In particular, a metaheuristic heterogeneous multiagent optimization framework is deployed. This approach embodies novel features of an intelligent system, similar to the human central nervous system, by exhibiting coordination and information sharing amongst several agents.

The algorithms developed are implemented on a model of an integrated gasification combined cycle (IGCC) plant with CO₂ capture. This plant includes significant mass and energy interactions with strong change in the coupling as the inputs are changed which renders it a prefect test case for the developed algorithm.

References

Friston, K. J., Harrison, L., & Penny, W. (2003). Dynamic causal modelling. NeuroImage, 1273-1302.

Halvorsen, I. J., Skogestad, S., Morud, J. C., & Alstad, V. (2003). Optimal selection of controlled variables. Industrial & Engineering Chemistry Research, 42, 3273-3284.

Jones, D., Bhattacharyya, D., Turton, R., & Zitney. (2014a). Plant-wide control system design: Primary controlled variable selection. Computers & Chemical Engineering, 220-234.

Jones, D., Bhattacharyya, D., Turton, R., & Zitney. (2014b). Plant-wide control system design: Secondary controlled variable selection. Computers & Chemical Engineering, 71, 253-262.