Papers by Liana Napalkova
International Journal of Aviation Management, 2021
integrated approach, supply chain simulation. The paper focuses on the development of simulation-... more integrated approach, supply chain simulation. The paper focuses on the development of simulation-based environment for multi-echelon cyclic planning and optimisation at the product maturity phase. It is based on integration of analytical and simulation techniques. Ana-lytical techniques are used to obtain initial planning deci-sions under conditions of stochastic demand and lead time. Simulation techniques extend these conditions to backlogging and capacity constraints. Simulation-based optimization is used to analyse and improve cyclical deci-sions received from the analytical model. Simulation en-vironment is built in the ProModel simulation software. Automatic generation of simulation model is provided by using the ProModel ActiveX technology and VBA pro-gramming language. An example of multi-echelon cyclic planning and optimisation using this environment is given.
Multi-objective simulation optimisation, evolutionary algorithms, hybrid algorithms Abstract – Th... more Multi-objective simulation optimisation, evolutionary algorithms, hybrid algorithms Abstract – The paper presents a taxonomic analysis of existing hybrid multi-objective evolutionary algorithms aimed at solving multi-objective simulation optimisation problems. For that, the properties of evolutionary algorithms and the requirements made to solving the problem considered are determined. Finally, a combination of the properties, which allows one to increase the approximation accuracy of the Pareto-optimal front at relatively low computational costs, is revealed.
error. NeuroFuzzy learning algorithm for simulation metamodelling is described in the paper. Here... more error. NeuroFuzzy learning algorithm for simulation metamodelling is described in the paper. Here, the simulation experimental data are used to train a fuzzy neural network-based simulation metamodel and to generate a set of relevant decision rules. Regression type model is applied to define the structure of the metamodel training set and essential decrease of an approximation error of simulation output data is received. The research results in the paper are illustrated with a range of experiments performed.
Abstract: The paper presents simulation-based methodology for analysis and optimisation of multi-... more Abstract: The paper presents simulation-based methodology for analysis and optimisation of multi-echelon supply chain planning policies over the product life cycle. It is aimed to analyse an efficiency of a specific planning policy at the product life cycle phases and to optimise the cyclic planning policy at the product maturity phase. Specific software prototypes and applications are described in the paper. The
This paper develops a multi-objective simulation-based genetic algorithm (MOSGA) for multi-echelo... more This paper develops a multi-objective simulation-based genetic algorithm (MOSGA) for multi-echelon supply chain cyclic planning and optimisation. The problem involves a search in high dimensional space with different ranges for decision variables scales, multiple objectives and problem specific constraints, such as power-of-two and nested/inverted-nested planning policies. In order to find the optimal solution, different parameters of genetic algorithm including the population sizing, crossover and mutation probabilities, selection and reproduction strategies and convergence criteria are investigated. For finding approximations of the Pareto optimal set, the non-dominated sorting approach is used. 1.
This paper develops a multi-objective simulation-based genetic algorithm (MOSGA) for multi- echel... more This paper develops a multi-objective simulation-based genetic algorithm (MOSGA) for multi- echelon supply chain cyclic planning and optimisation. The problem involves a search in high dimen- sional space with different ranges for decision variables scales, multiple objectives and problem specific constraints, such as power-of-two and nested/inverted-nested planning policies. In order to find the opti- mal solution, different parameters of genetic algorithm including the population sizing, crossover and mu- tation probabilities, selection and reproduction strategies and convergence criteria are investigated. For finding approximations of the Pareto optimal set, the non-dominated sorting approach is used.
The paper describes the algorithm, which is developed to solve scheduling tasks in Flexible Manuf... more The paper describes the algorithm, which is developed to solve scheduling tasks in Flexible Manufacturing Systems. The algorithm is a combination of Genetic Algorithm and Coloured Petri Nets. It is proposed to use Coloured Petri Nets to tackle the encoding problem in Genetic Algorithm. The objective is to minimize the total make-span subject to different constraints obtained in Flexible Manufacturing Systems.
The paper describes the algorithm, which is developed to solve scheduling tasks in Flexible Manuf... more The paper describes the algorithm, which is developed to solve scheduling tasks in Flexible Manufacturing Systems. The algorithm is a combination of Genetic Algorithm and Coloured Petri Nets. It is proposed to use Coloured Petri Nets to tackle the encoding problem in Genetic Algorithm. The objective is to minimize the total make-span subject to different constraints obtained in Flexible Manufacturing Systems.
The abundance of online media content requires highly scalable architectures to allow cross-media... more The abundance of online media content requires highly scalable architectures to allow cross-media monitoring. This paper presents an innovative big data-as-a-service platform for analysing large complex networks in order to enhance cross-media monitoring. In contrast to the existing media monitoring systems, the platform equips marketers with several distinctive features. First, while most of the systems perform quantitative exploratory analysis of social media, our platform applies graph analytics in order to reveal social interaction types, hidden patterns in the cross-media network and the information diffusion over time. Second, our platform integrates and implements distributed versions of graph analytics algorithms (Louvain, HITS and others) that can scale to a large volume of data. Third, the creation of cross-media graphs is triggered by user-defined queries that can be easily specified by marketers. Thus, end-users can build and analyse different graphs according to specifi...
This paper describes a two-phase simulation-based optimisation procedure that integrates the Gene... more This paper describes a two-phase simulation-based optimisation procedure that integrates the Genetic Algorithm and Response Surface-based Linear Search algorithm for developing optimal power-of-two replenishment policy in multi-echelon environment during the maturity phase of the life cycle of a product. The problem involves a search in high dimensional space with different ranges for decision variables scales, multiple objective functions and problem specific constraints, such as power-of-two and nested/invertednested planning policies. The paper provides illustrative example of the two-phase optimisation procedure applied to generic supply chain network.
This paper describes a system developed to assist in model-based training of minimally invasive, ... more This paper describes a system developed to assist in model-based training of minimally invasive, laparoscopic procedures. The key factor motivating the development of the device called CAST (Computer-Assisted Surgical Trainer) is the need to improve the state-of-the-art in teaching laparoscopy, and ultimately achieve better surgical outcomes. CAST's design concept and architecture is presented with its major elements that facilitate guided (both haptic and visual) execution of tasks, performance assessment, and comparative analysis of results. Both software and hardware models and implementations are given. The system, while currently intended for off-line, laboratory use, has an excellent potential for real-time assistive functions in the operating room.
The paper focuses on the development of simulationbased environment for multi-echelon cyclic plan... more The paper focuses on the development of simulationbased environment for multi-echelon cyclic planning and optimisation at the product maturity phase. It is based on integration of analytical and simulation techniques. Analytical techniques are used to obtain initial planning decisions under conditions of stochastic demand and lead time. Simulation techniques extend these conditions to backlogging and capacity constraints. Simulation-based optimization is used to analyse and improve cyclical decisions received from the analytical model. Simulation environment is built in the ProModel simulation software. Automatic generation of simulation model is provided by using the ProModel ActiveX technology and VBA programming language. An example of multi-echelon cyclic planning and optimisation using this environment is given.
Proceedings of the 20th European Modelling and Simulation Symposium, 2008
This paper describes a two-phase simulation-based optimisation procedure that integrates the Gene... more This paper describes a two-phase simulation-based optimisation procedure that integrates the Genetic Algorithm and Response Surface-based Linear Search algorithm for developing optimal power-of-two replenishment policy in multi-echelon environment during the maturity phase of the life cycle of a product. The problem involves a search in high dimensional space with different ranges for decision variables scales, multiple objective functions and problem specific constraints, such as power-of-two and nested/invertednested planning policies. The paper provides illustrative example of the two-phase optimisation procedure applied to generic supply chain network.
Proceedings of the European Modelling and Simulation Symposium, 2007
The paper focuses on the development of simulationbased environment for multi-echelon cyclic plan... more The paper focuses on the development of simulationbased environment for multi-echelon cyclic planning and optimisation at the product maturity phase. It is based on integration of analytical and simulation techniques. Analytical techniques are used to obtain initial planning decisions under conditions of stochastic demand and lead time. Simulation techniques extend these conditions to backlogging and capacity constraints. Simulation-based optimization is used to analyse and improve cyclical decisions received from the analytical model. Simulation environment is built in the ProModel simulation software. Automatic generation of simulation model is provided by using the ProModel ActiveX technology and VBA programming language. An example of multi-echelon cyclic planning and optimisation using this environment is given.
This paper develops a multi-objective simulation-based genetic algorithm (MOSGA) for multiechelon... more This paper develops a multi-objective simulation-based genetic algorithm (MOSGA) for multiechelon supply chain cyclic planning and optimisation. The problem involves a search in high dimensional space with different ranges for decision variables scales, multiple objectives and problem specific constraints, such as power-of-two and nested/inverted-nested planning policies. In order to find the optimal solution, different parameters of genetic algorithm including the population sizing, crossover and mutation probabilities, selection and reproduction strategies and convergence criteria are investigated. For finding approximations of the Pareto optimal set, the non-dominated sorting approach is used.
This paper focuses on the development of simulation-based environment for multi-echelon cyclic pl... more This paper focuses on the development of simulation-based environment for multi-echelon cyclic planning and optimisation in the product maturity phase. It is based on integration of analytical and simulation techniques. Analytical techniques are used to obtain initial planning decisions under conditions of stochastic demand and lead time, whereas simulation techniques extend these conditions to backlogging and capacity constraints. Simulation is used to analyse and improve cyclical decisions received from the analytical model. The proposed environment includes four components, such as database, process, optimisation and procedural one. Database component defines a supply chain network and its input parameters. Procedural component generates cyclic schedules using analytical calculus. Process component performs automatic generation of a supply chain simulation model and simulates cyclic schedules in multi-echelon environment while controlling inventory levels and estimating the performance measures. Optimisation component defines optimal cyclic schedule for each of the supply chain stages in order to minimize the sum of inventory holding, setup and ordering costs while satisfying customer service requirements defined by a target customer service level. The paper provides examples of different network type simulation models generated for multi-echelon cyclic planning and optimisation. The present research is funded by the ECLIPS Specific Targeted Research Project of the European Commission "Extended Collaborative Integrated Life Cycle Supply Chain Planning System".
NeuroFuzzy learning algorithm for simulation metamodelling is described in the paper. Here, the s... more NeuroFuzzy learning algorithm for simulation metamodelling is described in the paper. Here, the simulation experimental data are used to train a fuzzy neural network-based simulation metamodel and to generate a set of relevant decision rules. Regression type model is applied to define the structure of the metamodel training set and essential decrease of an approximation error of simulation output data is received. The research results in the paper are illustrated with a range of experiments performed.
Scientific Journal of Riga Technical University. Computer Sciences, 2009
This paper describes a system developed to assist in model-based training of minimally invasive, ... more This paper describes a system developed to assist in model-based training of minimally invasive, laparoscopic procedures. The key factor motivating the development of the device called CAST (Computer-Assisted Surgical Trainer) is the need to improve the state-of-the-art in teaching laparoscopy, and ultimately achieve better surgical outcomes. CAST's design concept and architecture is presented with its major elements that facilitate guided (both haptic and visual) execution of tasks, performance assessment, and comparative analysis of results. Both software and hardware models and implementations are given. The system, while currently intended for off-line, laboratory use, has an excellent potential for real-time assistive functions in the operating room.
Uploads
Papers by Liana Napalkova