Inverse hygric property determination based on dynamic measurements and swarm-intelligence optimisers

Abstract

To accelerate the hygric characterisation of porous building materials, dynamic flow and storage measurements in combination with inverse parameter estimation show a lot of promise. Therein though, the processing and interpretation of the experimental output can be challenging. This paper demonstrates the applicability of two swarmintelligence (SI) optimisers, i.e. the Particle Swarm Optimiser (PSO) and the Grey Wolf Optimiser (GWO), for determining the vapour resistance factor and the sorption isotherm of porous building materials. The methodology is presented for a fictitious dynamic vapour sorption experiment on a calcium silicate insulation sample. The identifiability of the unknown parameters and the reliability of the estimated properties is investigated via a profile likelihood (PL) analysis. By use of the proposed methodology, the measurement time required to determine the hygric properties is reduced strongly compared to standard techniques such as the steady-state cup and sorption tests, and at the same time the uncertainty propagated in the parameter estimation can be characterised. A close agreement with the target values is obtained.Though, to avoid an unreliable parameter estimation it is recommended to not limit the optimisation process to a single run. Furthermore, also the results obtained during the PL analysis can help improving the estimation. Finally, for the current case study, the SI optimisers are found to outperform the Genetic Algorithm (GA) and the Covariance Matrix Adaptation Evolution Strategy (CMA-ES).

Optimising Convolutional Neural Networks to Predict the Hygrothermal Performance of Building Components

“…However, this is usually a trial-and-error process, as there are no general guidelines. This paper hence proposes an approach to optimise these hyper-parameters, using the Grey-Wolf Optimisation (GWO) algorithm, as it was found competent for other applications [12,13]. This is applied to a one-dimensional (1D) brick wall, of which, the hygrothermal performance is evaluated for typical moisture damage patterns.…”

A critique on ‘A Boltzmann transformation method for investigation of water vapour transport in building materials’

“…The latter issue has not been addressed explicitly here, but the Boltzmann-Matano transformation requires a fixed boundary value, which cannot be guaranteed in an hygroscopic adsorption test with a non-zero surface vapour transfer resistance. Given these concerns, this correction of the Pavlik et al (2011) approach cannot be encouraged, and instead inverse hygric property determination (Vereecken et al, 2018) should (perhaps) be preferred. In this method, the vapour storage and vapour transport properties are determined simultaneously from short-term dynamic experiments, and their processing moreover allows explicitly considering the impacts of the surface vapour transfer resistance.…”

Centrifuge experiments for the determination of the moisture storage and transport properties in the overhygroscopic range

“…For noise-free data, it should be possible to infer the permeability. Indeed, when creating artificial data for a similar centrifuge experiment (see [5] for the principle applied on moisture transport in the hygroscopic range) on a brick sample with known moisture storage and transport properties, both the moisture retention and the permeability curve could be accurately inferred (not shown).…”