Adam Kiczko

Differential Evolution algorithm with Separated Groups for multi-dimensional optimization problems

The classical Differential Evolution (DE) algorithm, one of population-based Evolutionary Computation methods, proved to be a successful approach for relatively simple problems, but does not perform well for difficult multi-dimensional non-convex functions. A number of significant modifications of DE have been proposed in recent years, including very few approaches referring to the idea of distributed Evolutionary Algorithms. The present paper presents a new algorithm to improve optimization performance, namely DE with Separated Groups (DE-SG), which distributes population into small groups, defines rules of exchange of information and individuals between the groups and uses two different strategies to keep balance between exploration and exploitation capabilities. The performance of DE-SG is compared to that of eight algorithms belonging to the class of Evolutionary Strategies (Covariance Matrix Adaptation ES), Particle Swarm Optimization (Comprehensive Learning PSO and Efficient Population Utilization Strategy PSO), Differential Evolution (Distributed DE with explorative-exploitative population families, Self-adaptive DE, DE with global and local neighbours and Grouping Differential Evolution) and multi-algorithms (AMALGAM). The comparison is carried out for a set of 10-, 30- and 50-dimensional rotated test problems of varying difficulty, including 10- and 30-dimensional composition functions from CEC2005. Although slow for simple functions, the proposed DE-SG algorithm achieves a great success rate for more difficult 30- and 50-dimensional problems.

Corrigendum to: “Differential evolution algorithm with separated groups for multi-dimensional optimization problems” [Eur. J. Oper. Res. 216 (2012) 33–46]

0377-2217/

Stochastic transfer function model applied to combined reservoir management and flow routing

see front matter 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.ejor.2011.12.043 DOI of original article: 10.1016/j.ejor.2011.07.038 ⇑ Corresponding author. Tel.: +48 6915 858. E-mail address: adampp@igf.edu.pl (A.P. Piotrowski). The random values rand(0,1) and RandNorm(0,1) should be generated only once for each individual in each generation. The strategy A is composed of two mutation schemes, but only one of them is used for particular parent.

Experimental Investigation of Kinetic Energy and Momentum Coefficients in Regular Channels with Stiff and Flexible Elements Simulating Submerged Vegetation

Abstract The paper presents an application of a Stochastic Transfer Function (STF) approach and a nonlinear transformation of model variables to combined reservoir management and flow routing on the Upper Narew River, northeast Poland. The management objective is to reach the required flow conditions in the reaches of an ecologically valuable river. A 1-D distributed flow routing model was designed for the study. However, both optimization methods and reservoir management analysis require numerous model realizations which are computationally very expensive. A much more efficient solution consists of the application of a simplified STF simulator of river flow, which is calibrated on historical data and simulated flows via a distributed model for those parts of the river where observations are not available. The model is stochastic, enabling the derivation of a simulator prediction uncertainty in a straightforward manner. The optimal control policy obtained is tested on a fully distributed model. Obviously, the simulators predictive uncertainty is larger than the uncertainty of the original distributed model. Therefore, in this application the original model is used to derive probability maps of inundation extent at high- and low-flow conditions under optimized reservoir discharges, following the Generalized Likelihood Uncertainty Estimation procedure. The main advantage of using the simulator lies in the much shorter computation times required by the water management optimization problem, which would be very difficult to solve with a distributed model used for flow routing. Additionally, the simulator may be used for the estimation of probability maps of inundation extent in the case where a distributed flow routing model cannot be run in a Monte Carlo set-up. Citation Romanowicz, R. J., Kiczko, A. & Napiórkowski, J. J. (2010) Stochastic transfer function model applied to combined reservoir management and flow routing. Hydrol. Sci. J. 55(1), 27–40.

Uncertainty of Deardorff’s soil moisture model based on continuous TDR measurements for sandy loam soil

The paper addresses the problem of determination of the energy and momentum coefficients for flows through a partly vegetated channel. These coefficients are applied to express the fluid kinetic energy and momentum equations as functions of a mean velocity. The study is based on laboratory measurements of water velocity distributions in a straight rectangular flume with stiff and flexible stems and plastic imitations of the Canadian waterweed. The coefficients were established for the vegetation layer, surface layer and the whole flow area. The results indicate that the energy and momentum coefficients increase significantly with water depth and the number of stems per unit channel area. New regression relationships for both coefficients are given.

Experimental and numerical investigation of non-submerged flow under a sluice gate

Abstract Knowledge on soil moisture is indispensable for a range of hydrological models, since it exerts a considerable influence on runoff conditions. Proper tools are nowadays applied in order to gain in-sight into soil moisture status, especially of uppermost soil layers, which are prone to weather changes and land use practices. In order to establish relationships between meteorological conditions and topsoil moisture, a simple model would be required, characterized by low computational effort, simple structure and low number of identified and calibrated parameters. We demonstrated, that existing model for shallow soils, considering mass exchange between two layers (the upper and the lower), as well as with the atmosphere and subsoil, worked well for sandy loam with deep ground water table in Warsaw conurbation. GLUE (Generalized Likelihood Uncertainty Estimation) linked with GSA (Global Sensitivity Analysis) provided for final determination of parameter values and model confidence ranges. Including the uncertainty in a model structure, caused that the median soil moisture solution of the GLUE was shifted from the one optimal in deterministic sense. From the point of view of practical model application, the main shortcoming were the underestimated water exchange rates between the lower soil layer (ranging from the depth of 0.1 to 0.2 m below ground level) and subsoil. General model quality was found to be satisfactory and promising for its utilization for establishing measures to regain retention in urbanized conditions.

Too wet and too dry? Uncertainty of DEM as a potential source of significant errors in a model-based water level assessment in riparian and mire ecosystems

Abstract The problem of sluice gate flow is analyzed using two models: a simplified one, derived according to the concept of the Potential Field (PF), and a more complex form, based on the Reynolds Average Navier-Stokes (RANS) equations. The numerical solution is compared with experimental data, including measurements performed by authors and results acquired from literature. Despite its simplicity, the PF model provides a satisfactory agreement with the measurements. The slightly worse performance of the RANS model comes from an overestimation of energy losses.

Turbulence intensity and spatial scales of turbulence after hydraulic jump over scour hole in rectangular channel

Modelling groundwater depths in floodplains and peatlands remains a basic approach to assessing hydrological conditions of habitats. Groundwater flow models used to compute groundwater heads are known for their uncertainties, and the calibration of these models and the uncertainty assessments of parameters remain fundamental steps in providing reliable data. However, the elevation data used to determine the geometry of model domains are frequently considered deterministic and hence are seldom considered a source of uncertainty in model-based groundwater level estimations. Knowing that even the cutting-edge laser-scanning-based digital elevation models have errors due to vegetation effects and scanning procedure failures, we provide an assessment of uncertainty of water level estimations that remain basic data for wetland ecosystem assessment and management. We found that the uncertainty of the digital elevation model (DEM) significantly influenced the results of the assessment of the habitat’s hydrological conditions expressed as groundwater depths. In extreme cases, although the average habitat suitability index (HSI) assessed in a deterministic manner was defined as ‘unsuitable’, in a probabilistic approach (grid-cell-scale estimation), it reached a value of 40% probability, signifying ‘optimum’ or ‘tolerant’. For the 24 habitats analysed, we revealed vast differences between HSI scores calculated for individual grid cells of the model and HSI scores computed as average values from the set of grid cells located within the habitat patches. We conclude that groundwater-modelling-based decision support approaches to wetland assessment can result in incorrect management if the quality of DEM has not been addressed in studies referring to groundwater depths.

Modeling of water flow in multi-channel river system in the Narew National Park

Abstract The study presents experimental investigations of spatial turbulence intensity and scales of turbulent eddies (macroeddies) in a rectangular channel and the impact of the hydraulic jump on their vertical and streamwise distributions over a flat and scoured bed. The results of four tests and two different discharge rates are presented. Intensive mixing caused by the hydraulic jump has an impact on the instantaneous velocity, turbulence intensity and sizes of macroeddies, as well as their vertical and longitudinal distributions along the channel. The largest differences in turbulence characteristics were reported directly after the hydraulic jump, above the eroded bed. The interaction between the stream of the increased turbulence and the bed is a direct cause of formation of scour downstream water structures, which has a great effect on overall flow characteristics. The scour hole that arose downstream the jump moderated, in a small degree, the turbulence intensity at its end. Just next to the hydraulic jump only the small longitudinal relative sizes of macroeddies were present, while at the end of the analyzed reach, downstream of the scour, the relative scale reached around 1.5 depth of the stream.

Turbulent intensity and scales of turbulence after hydraulic jump in rectangular channel

Abstract Modeling of water flow in multi-channel river system in the Narew National Park. Anastomosing rivers constitute a rare example of multi-channel systems, which used to be very common before the agricultural and industrial development. Presently few of them remain worldwide and the only example in Poland is the Upper River Narew within Narew National Park. Although hydraulic modeling using one-dimensional models is commonly used to describe water flow in rivers, for multi-channel rivers problem is more complicated. For this type of rivers it is expected that the feedback between process of plants growth (expressed by Manning’s coefficient) and distribution of flow in anabranches is high. However, assignment procedure on roughness coefficients in splitting and rejoining channels is laborious and difficult. Therefore, for efficient water flow modeling in multi-channel systems a stand-alone hydraulic model equipped with automatic optimization procedure was developed. Optimization and validation stages, based on field measurements data of discharge and water levels, indicated that the model accurately simulates water flow in multi-channel system.