Populaation monimuotoisuuden mittaaminen liukulukukoodatuissa evoluutioalgoritmeissa

Diplomityössä esitetään menetelmä populaation monimuotoisuuden mittaamiseen liukulukukoodatuissa evoluutioalgoritmeissa, ja tarkastellaan kokeellisesti sen toimintaa. Evoluutioalgoritmit ovat populaatiopohjaisia menetelmiä, joilla pyritään ratkaisemaan optimointiongelmia. Evoluutioalgoritmeissa populaation monimuotoisuuden hallinta on välttämätöntä, jotta suoritettu haku olisi riittävän luotettavaa ja toisaalta riittävän nopeaa. Monimuotoisuuden mittaaminen on erityisen tarpeellista tutkittaessa evoluutioalgoritmien dynaamista käyttäytymistä. Työssä tarkastellaan haku- ja tavoitefunktioavaruuden monimuotoisuuden mittaamista. Toistaiseksi ei ole ollut olemassa täysin tyydyttäviä monimuotoisuuden mittareita, ja työn tavoitteena on kehittää yleiskäyttöinen menetelmä liukulukukoodattujen evoluutioalgoritmien suhteellisen ja absoluuttisen monimuotoisuuden mittaamiseen hakuavaruudessa. Kehitettyjen mittareiden toimintaa ja käyttökelpoisuutta tarkastellaan kokeellisesti ratkaisemalla optimointiongelmia differentiaalievoluutioalgoritmilla. Toteutettujen mittareiden toiminta perustuu keskihajontojen laskemiseen populaatiosta. Keskihajonnoille suoritetaan skaalaus, joko alkupopulaation tai nykyisen populaation suhteen, riippuen lasketaanko absoluuttista vai suhteellista monimuotoisuutta. Kokeellisessa tarkastelussa havaittiin kehitetyt mittarit toimiviksi ja käyttökelpoisiksi. Tavoitefunktion venyttäminen koordinaattiakseleiden suunnassa ei vaikuta mittarin toimintaan. Myöskään tavoitefunktion kiertäminen koordinaatistossa ei vaikuta mittareiden tuloksiin. Esitetyn menetelmän aikakompleksisuus riippuu lineaarisesti populaation koosta, ja mittarin toiminta on siten nopeaa suuriakin populaatioita käytettäessä. Suhteellinen monimuotoisuus antaa vertailukelpoisia tuloksia riippumatta parametrien lukumäärästä tai populaation koosta.

Trust region versus line search for computing the optical flow

We consider the numerical treatment of the optical flow problem by evaluating the performance of the trust region method versus the line search method. To the best of our knowledge, the trust region method is studied here for the first time for variational optical flow computation. Four different optical flow models are used to test the performance of the proposed algorithm combining linear and nonlinear data terms with quadratic and TV regularization. We show that trust region often performs better than line search; especially in the presence of non-linearity and non-convexity in the model.

Estudo do estado fundamental de aglomerados de silício via redes neurais

We introduce a global optimization method based on the cooperation between an Artificial Neural Net (ANN) and Genetic Algorithm (GA). We have used ANN to select the initial population for the GA. We have tested the new method to predict the ground-state geometry of silicon clusters. We have described the clusters as a piling of plane structures. We have trained three ANN architectures and compared their results with those of pure GA. ANN strongly reduces the total computational time. For Si10, it gained a factor of 5 in search speed. This method can be easily extended to other optimization problems.

Analyzing Carbon Trade Permits in Optimization Framework

The threats caused by global warming motivate different stake holders to deal with and control them. This Master's thesis focuses on analyzing carbon trade permits in optimization framework. The studied model determines optimal emission and uncertainty levels which minimize the total cost. Research questions are formulated and answered by using different optimization tools. The model is developed and calibrated by using available consistent data in the area of carbon emission technology and control. Data and some basic modeling assumptions were extracted from reports and existing literatures. The data collected from the countries in the Kyoto treaty are used to estimate the cost functions. Theory and methods of constrained optimization are briefly presented. A two-level optimization problem (individual and between the parties) is analyzed by using several optimization methods. The combined cost optimization between the parties leads into multivariate model and calls for advanced techniques. Lagrangian, Sequential Quadratic Programming and Differential Evolution (DE) algorithm are referred to. The role of inherent measurement uncertainty in the monitoring of emissions is discussed. We briefly investigate an approach where emission uncertainty would be described in stochastic framework. MATLAB software has been used to provide visualizations including the relationship between decision variables and objective function values. Interpretations in the context of carbon trading were briefly presented. Suggestions for future work are given in stochastic modeling, emission trading and coupled analysis of energy prices and carbon permits.

DIRECT INFUSION ESI-MS APPLIED IN THE DETECTION OF BYPRODUCTS DUE TO REDUCTIVE DEGRADATION OF ACETAMIPRID BY ZERO-VALENT IRON

This study investigated the reductive degradation of acetamiprid (5 mg L-1) in aqueous medium (at pH 2.0) induced by zero-valent iron (50 mg). The process was monitored using high-performance liquid chromatography (HPLC) to determine the degradation rate as a function of reaction time, and direct infusion electrospray ionization mass spectrometry (DI-ESI-MS) to search for (and potentially characterize) any possible byproducts formed during degradation. The results obtained via HPLC showed that after 60 min, the degradation of the substrate reached nearly 100% in an acidic medium, whereas the mineralization rate (as determined by total organic carbon measurements) was as low as 3%. Data obtained by DI-ESI-MS showed that byproducts were formed mainly by insertions of hydrogen atoms into the nitrile, imine, and pyridine ring moieties, in addition to the observation of chlorine substitution by hydrogen replacement (hydrodechlorination) reactions.

Learning Robot Environment through Simulation

Traditionally simulators have been used extensively in robotics to develop robotic systems without the need to build expensive hardware. However, simulators can be also be used as a “memory”for a robot. This allows the robot to try out actions in simulation before executing them for real. The key obstacle to this approach is an uncertainty of knowledge about the environment. The goal of the Master’s Thesis work was to develop a method, which allows updating the simulation model based on actual measurements to achieve a success of the planned task. OpenRAVE was chosen as an experimental simulation environment on planning,trial and update stages. Steepest Descent algorithm in conjunction with Golden Section search procedure form the principle part of optimization process. During experiments, the properties of the proposed method, such as sensitivity to different parameters, including gradient and error function, were examined. The limitations of the approach were established, based on analyzing the regions of convergence.

Application of simulated annealing in simulation and optimization of drying process of Zea mays malt

Kinetic simulation and drying process optimization of corn malt by Simulated Annealing (SA) for estimation of temperature and time parameters in order to preserve maximum amylase activity in the obtained product are presented here. Germinated corn seeds were dried at 54-76 °C in a convective dryer, with occasional measurement of moisture content and enzymatic activity. The experimental data obtained were submitted to modeling. Simulation and optimization of the drying process were made by using the SA method, a randomized improvement algorithm, analogous to the simulated annealing process. Results showed that seeds were best dried between 3h and 5h. Among the models used in this work, the kinetic model of water diffusion into corn seeds showed the best fitting. Drying temperature and time showed a square influence on the enzymatic activity. Optimization through SA showed the best condition at 54 ºC and between 5.6h and 6.4h of drying. Values of specific activity in the corn malt were found between 5.26±0.06 SKB/mg and 15.69±0,10% of remaining moisture.

The Possibilities and Dosimetric Limitations of MLC-Based Intensity-Modulated Radiotherapy Delivery and Optimization Techniques

The use of intensity-modulated radiotherapy (IMRT) has increased extensively in the modern radiotherapy (RT) treatments over the past two decades. Radiation dose distributions can be delivered with higher conformality with IMRT when compared to the conventional 3D-conformal radiotherapy (3D-CRT). Higher conformality and target coverage increases the probability of tumour control and decreases the normal tissue complications. The primary goal of this work is to improve and evaluate the accuracy, efficiency and delivery techniques of RT treatments by using IMRT. This study evaluated the dosimetric limitations and possibilities of IMRT in small (treatments of head-and-neck, prostate and lung cancer) and large volumes (primitive neuroectodermal tumours). The dose coverage of target volumes and the sparing of critical organs were increased with IMRT when compared to 3D-CRT. The developed split field IMRT technique was found to be safe and accurate method in craniospinal irradiations. By using IMRT in simultaneous integrated boosting of biologically defined target volumes of localized prostate cancer high doses were achievable with only small increase in the treatment complexity. Biological plan optimization increased the probability of uncomplicated control on average by 28% when compared to standard IMRT delivery. Unfortunately IMRT carries also some drawbacks. In IMRT the beam modulation is realized by splitting a large radiation field to small apertures. The smaller the beam apertures are the larger the rebuild-up and rebuild-down effects are at the tissue interfaces. The limitations to use IMRT with small apertures in the treatments of small lung tumours were investigated with dosimetric film measurements. The results confirmed that the peripheral doses of the small lung tumours were decreased as the effective field size was decreased. The studied calculation algorithms were not able to model the dose deficiency of the tumours accurately. The use of small sliding window apertures of 2 mm and 4 mm decreased the tumour peripheral dose by 6% when compared to 3D-CRT treatment plan. A direct aperture based optimization (DABO) technique was examined as a solution to decrease the treatment complexity. The DABO IMRT technique was able to achieve treatment plans equivalent with the conventional IMRT fluence based optimization techniques in the concave head-and-neck target volumes. With DABO the effective field sizes were increased and the number of MUs was reduced with a factor of two. The optimality of a treatment plan and the therapeutic ratio can be further enhanced by using dose painting based on regional radiosensitivities imaged with functional imaging methods.

Qualitative Characteristics and Quantitative Measures of Solution's Reliability in Discrete Optimization: Traditional Analytical Approaches, Innovative Computational Methods and Applicability

The purpose of this thesis is twofold. The first and major part is devoted to sensitivity analysis of various discrete optimization problems while the second part addresses methods applied for calculating measures of solution stability and solving multicriteria discrete optimization problems. Despite numerous approaches to stability analysis of discrete optimization problems two major directions can be single out: quantitative and qualitative. Qualitative sensitivity analysis is conducted for multicriteria discrete optimization problems with minisum, minimax and minimin partial criteria. The main results obtained here are necessary and sufficient conditions for different stability types of optimal solutions (or a set of optimal solutions) of the considered problems. Within the framework of quantitative direction various measures of solution stability are investigated. A formula for a quantitative characteristic called stability radius is obtained for the generalized equilibrium situation invariant to changes of game parameters in the case of the H¨older metric. Quality of the problem solution can also be described in terms of robustness analysis. In this work the concepts of accuracy and robustness tolerances are presented for a strategic game with a finite number of players where initial coefficients (costs) of linear payoff functions are subject to perturbations. Investigation of stability radius also aims to devise methods for its calculation. A new metaheuristic approach is derived for calculation of stability radius of an optimal solution to the shortest path problem. The main advantage of the developed method is that it can be potentially applicable for calculating stability radii of NP-hard problems. The last chapter of the thesis focuses on deriving innovative methods based on interactive optimization approach for solving multicriteria combinatorial optimization problems. The key idea of the proposed approach is to utilize a parameterized achievement scalarizing function for solution calculation and to direct interactive procedure by changing weighting coefficients of this function. In order to illustrate the introduced ideas a decision making process is simulated for three objective median location problem. The concepts, models, and ideas collected and analyzed in this thesis create a good and relevant grounds for developing more complicated and integrated models of postoptimal analysis and solving the most computationally challenging problems related to it.

On Innovative Search: the use of internal and external sources of innovation among Finnish innovators

This dissertation explores the use of internal and external sources of knowledge in modern innovation processes. It builds on a framework that combines theories such as a behavioural theory of the firm, the evolutionary theory of economic change, and modern approaches to strategic management. It follows the recent increase in innovation research focusing on the firm-level examination of innovative activities instead of traditional industry-level determinants. The innovation process is seen as a problem- and slack- driven search process, which can take several directions in terms of organizational boundaries in the pursuit of new knowledge and other resources. It thus draws on recent models of technological change, according to which firms nowadays should build their innovative activities on both internal and external sources of innovation rather than relying solely on internal resources. Four different research questions are addressed, all of which are empirically investigated via a rich dataset covering Finnish innovators collected by Statistics Finland. Firstly, the study examines how the nature of problems shapes the direction of any search for new knowledge. In general it demonstrates that the nature of the problem does affect the direction of the search, although under resource constraints firms tend to use external rather than internal sources of knowledge. At the same time, it shows that those firms that are constrained in terms of finance seem to search both internally and externally. Secondly, the dissertation investigates the relationships between different kinds of internal and external sources of knowledge in an attempt to find out where firms should direct their search in order to exploit the potential of a distributed innovation process. The concept of complementarities is applied in this context. The third research question concerns how the use of external knowledge sources – openness to external knowledge – influences the financial performance of firms. Given the many advantages of openness presented in the current literature, the focus is on how it shapes profitability. The results reveal a curvilinear relationship between profitability and openness (taking an inverted U-shape), the implication being that it pays to be open up to a certain point, but being too open to external sources may be detrimental to financial performance. Finally, the dissertation addresses some challenges in CISbased innovation research that have received relatively little attention in prior studies. The general aim is to underline the fact that comprehensive understanding of the complex process of technological change requires the constant development of methodological approaches (in terms of data and measures, for example). All the empirical analyses included in the dissertation are based on the Finnish CIS (Finnish Innovation Survey 1998-2000).

Search and Paper SKU Analysis of Recipes Meeting EUPS Standard

The theoretical research of the study concentrated on finding theoretical frameworks to optimize the amount of needed stock keeping units (SKUs) in manufacturing industry. The goal was to find ways for a company to acquire an optimal collection of stock keeping units needed for manufacturing needed amount of end products. The research follows constructive research approach leaning towards practical problem solving. In the empirical part of this study, a recipe search tool was developed to an existing database used in the target company. The purpose of the tools was to find all the recipes meeting the EUPS performance standard and put the recipes in a ranking order using the data available in the database. The ranking of the recipes was formed from the combination of the performance measures and price of the recipes. In addition, the tool researched what kind of paper SKUs were needed to manufacture the best performing recipes. The tool developed during this process meets the requirements. It eases and makes it much faster to search for all the recipes meeting the EUPS standard. Furthermore, many future development possibilities for the tool were discovered while writing the thesis.

Stochastic approximation methods in stochastic optimization

Stochastic approximation methods for stochastic optimization are considered. Reviewed the main methods of stochastic approximation: stochastic quasi-gradient algorithm, Kiefer-Wolfowitz algorithm and adaptive rules for them, simultaneous perturbation stochastic approximation (SPSA) algorithm. Suggested the model and the solution of the retailer's profit optimization problem and considered an application of the SQG-algorithm for the optimization problems with objective functions given in the form of ordinary differential equation.

Towards sustainable Iron- and steelmaking with economic optimization

The iron and steelmaking industry is among the major contributors to the anthropogenic emissions of carbon dioxide in the world. The rising levels of CO2 in the atmosphere and the global concern about the greenhouse effect and climate change have brought about considerable investigations on how to reduce the energy intensity and CO2 emissions of this industrial sector. In this thesis the problem is tackled by mathematical modeling and optimization using three different approaches. The possibility to use biomass in the integrated steel plant, particularly as an auxiliary reductant in the blast furnace, is investigated. By pre-processing the biomass its heating value and carbon content can be increased at the same time as the oxygen content is decreased. As the compression strength of the preprocessed biomass is lower than that of coke, it is not suitable for replacing a major part of the coke in the blast furnace burden. Therefore the biomass is assumed to be injected at the tuyere level of the blast furnace. Carbon capture and storage is, nowadays, mostly associated with power plants but it can also be used to reduce the CO2 emissions of an integrated steel plant. In the case of a blast furnace, the effect of CCS can be further increased by recycling the carbon dioxide stripped top gas back into the process. However, this affects the economy of the integrated steel plant, as the amount of top gases available, e.g., for power and heat production is decreased. High quality raw materials are a prerequisite for smooth blast furnace operation. High quality coal is especially needed to produce coke with sufficient properties to ensure proper gas permeability and smooth burden descent. Lower quality coals as well as natural gas, which some countries have in great volumes, can be utilized with various direct and smelting reduction processes. The DRI produced with a direct reduction process can be utilized as a feed material for blast furnace, basic oxygen furnace or electric arc furnace. The liquid hot metal from a smelting reduction process can in turn be used in basic oxygen furnace or electric arc furnace. The unit sizes and investment costs of an alternative ironmaking process are also lower than those of a blast furnace. In this study, the economy of an integrated steel plant is investigated by simulation and optimization. The studied system consists of linearly described unit processes from coke plant to steel making units, with a more detailed thermodynamical model of the blast furnace. The results from the blast furnace operation with biomass injection revealed the importance of proper pre-processing of the raw biomass as the composition of the biomass as well as the heating value and the yield are all affected by the pyrolysis temperature. As for recycling of CO2 stripped blast furnace top gas, substantial reductions in the emission rates are achieved if the stripped CO2 can be stored. However, the optimal recycling degree together with other operation conditions is heavily dependent on the cost structure of CO2 emissions and stripping/storage. The economical feasibility related to the use of DRI in the blast furnace depends on the price ratio between the DRI pellets and the BF pellets. The high amount of energy needed in the rotary hearth furnace to reduce the iron ore leads to increased CO2 emissions.

Stiffness based trajectory planning and feedforward based vibration suppression control of parallel robot machines

The dissertation proposes two control strategies, which include the trajectory planning and vibration suppression, for a kinematic redundant serial-parallel robot machine, with the aim of attaining the satisfactory machining performance. For a given prescribed trajectory of the robot's end-effector in the Cartesian space, a set of trajectories in the robot's joint space are generated based on the best stiffness performance of the robot along the prescribed trajectory. To construct the required system-wide analytical stiffness model for the serial-parallel robot machine, a variant of the virtual joint method (VJM) is proposed in the dissertation. The modified method is an evolution of Gosselin's lumped model that can account for the deformations of a flexible link in more directions. The effectiveness of this VJM variant is validated by comparing the computed stiffness results of a flexible link with the those of a matrix structural analysis (MSA) method. The comparison shows that the numerical results from both methods on an individual flexible beam are almost identical, which, in some sense, provides mutual validation. The most prominent advantage of the presented VJM variant compared with the MSA method is that it can be applied in a flexible structure system with complicated kinematics formed in terms of flexible serial links and joints. Moreover, by combining the VJM variant and the virtual work principle, a systemwide analytical stiffness model can be easily obtained for mechanisms with both serial kinematics and parallel kinematics. In the dissertation, a system-wide stiffness model of a kinematic redundant serial-parallel robot machine is constructed based on integration of the VJM variant and the virtual work principle. Numerical results of its stiffness performance are reported. For a kinematic redundant robot, to generate a set of feasible joints' trajectories for a prescribed trajectory of its end-effector, its system-wide stiffness performance is taken as the constraint in the joints trajectory planning in the dissertation. For a prescribed location of the end-effector, the robot permits an infinite number of inverse solutions, which consequently yields infinite kinds of stiffness performance. Therefore, a differential evolution (DE) algorithm in which the positions of redundant joints in the kinematics are taken as input variables was employed to search for the best stiffness performance of the robot. Numerical results of the generated joint trajectories are given for a kinematic redundant serial-parallel robot machine, IWR (Intersector Welding/Cutting Robot), when a particular trajectory of its end-effector has been prescribed. The numerical results show that the joint trajectories generated based on the stiffness optimization are feasible for realization in the control system since they are acceptably smooth. The results imply that the stiffness performance of the robot machine deviates smoothly with respect to the kinematic configuration in the adjacent domain of its best stiffness performance. To suppress the vibration of the robot machine due to varying cutting force during the machining process, this dissertation proposed a feedforward control strategy, which is constructed based on the derived inverse dynamics model of target system. The effectiveness of applying such a feedforward control in the vibration suppression has been validated in a parallel manipulator in the software environment. The experimental study of such a feedforward control has also been included in the dissertation. The difficulties of modelling the actual system due to the unknown components in its dynamics is noticed. As a solution, a back propagation (BP) neural network is proposed for identification of the unknown components of the dynamics model of the target system. To train such a BP neural network, a modified Levenberg-Marquardt algorithm that can utilize an experimental input-output data set of the entire dynamic system is introduced in the dissertation. Validation of the BP neural network and the modified Levenberg- Marquardt algorithm is done, respectively, by a sinusoidal output approximation, a second order system parameters estimation, and a friction model estimation of a parallel manipulator, which represent three different application aspects of this method.

Lightweight, liquid-cooled, direct-drive generator for highpower wind turbines: motivation, concept, and performance

Thesis: A liquid-cooled, direct-drive, permanent-magnet, synchronous generator with helical, double-layer, non-overlapping windings formed from a copper conductor with a coaxial internal coolant conduit offers an excellent combination of attributes to reliably provide economic wind power for the coming generation of wind turbines with power ratings between 5 and 20MW. A generator based on the liquid-cooled architecture proposed here will be reliable and cost effective. Its smaller size and mass will reduce build, transport, and installation costs. Summary: Converting wind energy into electricity and transmitting it to an electrical power grid to supply consumers is a relatively new and rapidly developing method of electricity generation. In the most recent decade, the increase in wind energy’s share of overall energy production has been remarkable. Thousands of land-based and offshore wind turbines have been commissioned around the globe, and thousands more are being planned. The technologies have evolved rapidly and are continuing to evolve, and wind turbine sizes and power ratings are continually increasing. Many of the newer wind turbine designs feature drivetrains based on Direct-Drive, Permanent-Magnet, Synchronous Generators (DD-PMSGs). Being low-speed high-torque machines, the diameters of air-cooled DD-PMSGs become very large to generate higher levels of power. The largest direct-drive wind turbine generator in operation today, rated just below 8MW, is 12m in diameter and approximately 220 tonne. To generate higher powers, traditional DD-PMSGs would need to become extraordinarily large. A 15MW air-cooled direct-drive generator would be of colossal size and tremendous mass and no longer economically viable. One alternative to increasing diameter is instead to increase torque density. In a permanent magnet machine, this is best done by increasing the linear current density of the stator windings. However, greater linear current density results in more Joule heating, and the additional heat cannot be removed practically using a traditional air-cooling approach. Direct liquid cooling is more effective, and when applied directly to the stator windings, higher linear current densities can be sustained leading to substantial increases in torque density. The higher torque density, in turn, makes possible significant reductions in DD-PMSG size. Over the past five years, a multidisciplinary team of researchers has applied a holistic approach to explore the application of liquid cooling to permanent-magnet wind turbine generator design. The approach has considered wind energy markets and the economics of wind power, system reliability, electromagnetic behaviors and design, thermal design and performance, mechanical architecture and behaviors, and the performance modeling of installed wind turbines. This dissertation is based on seven publications that chronicle the work. The primary outcomes are the proposal of a novel generator architecture, a multidisciplinary set of analyses to predict the behaviors, and experimentation to demonstrate some of the key principles and validate the analyses. The proposed generator concept is a direct-drive, surface-magnet, synchronous generator with fractional-slot, duplex-helical, double-layer, non-overlapping windings formed from a copper conductor with a coaxial internal coolant conduit to accommodate liquid coolant flow. The novel liquid-cooling architecture is referred to as LC DD-PMSG. The first of the seven publications summarized in this dissertation discusses the technological and economic benefits and limitations of DD-PMSGs as applied to wind energy. The second publication addresses the long-term reliability of the proposed LC DD-PMSG design. Publication 3 examines the machine’s electromagnetic design, and Publication 4 introduces an optimization tool developed to quickly define basic machine parameters. The static and harmonic behaviors of the stator and rotor wheel structures are the subject of Publication 5. And finally, Publications 6 and 7 examine steady-state and transient thermal behaviors. There have been a number of ancillary concrete outcomes associated with the work including the following. X Intellectual Property (IP) for direct liquid cooling of stator windings via an embedded coaxial coolant conduit, IP for a lightweight wheel structure for lowspeed, high-torque electrical machinery, and IP for numerous other details of the LC DD-PMSG design X Analytical demonstrations of the equivalent reliability of the LC DD-PMSG; validated electromagnetic, thermal, structural, and dynamic prediction models; and an analytical demonstration of the superior partial load efficiency and annual energy output of an LC DD-PMSG design X A set of LC DD-PMSG design guidelines and an analytical tool to establish optimal geometries quickly and early on X Proposed 8 MW LC DD-PMSG concepts for both inner and outer rotor configurations Furthermore, three technologies introduced could be relevant across a broader spectrum of applications. 1) The cost optimization methodology developed as part of this work could be further improved to produce a simple tool to establish base geometries for various electromagnetic machine types. 2) The layered sheet-steel element construction technology used for the LC DD-PMSG stator and rotor wheel structures has potential for a wide range of applications. And finally, 3) the direct liquid-cooling technology could be beneficial in higher speed electromotive applications such as vehicular electric drives.