Rashad Rafik Aliev

Fuzzy-genetic approach to aggregate production-distribution planning in supply chain management

Aggregate production-distribution planning (APDP) is one of the most important activities in supply chain management (SCM). When solving the problem of APDP, we are usually faced with uncertain market demands and capacities in production environment, imprecise process times, and other factors introducing inherent uncertainty to the solution. Using deterministic and stochastic models in such conditions may not lead to fully satisfactory results. Using fuzzy models allows us to remove this drawback. It also facilitates the inclusion of expert knowledge. However, the majority of existing fuzzy models deal only with separate aggregate production planning without taking into account the interrelated nature of production and distribution systems. This limited approach often leads to inadequate results. An integration of the two interconnected processes within a single production-distribution model would allow better planning and management. Due to the need for a joint general strategic plan for production and distribution and vague planning data, in this paper we develop a fuzzy integrated multi-period and multi-product production and distribution model in supply chain. The model is formulated in terms of fuzzy programming and the solution is provided by genetic optimization (genetic algorithm). The use of the interactive aggregate production-distribution planning procedure developed on the basis of the proposed fuzzy integrated model with fuzzy objective function and soft constraints allows sound trade-off between the maximization of profit and fillrate. The experimental results demonstrate high efficiency of the proposed method.

Evolutionary algorithm-based learning of fuzzy neural networks. Part 2: Recurrent fuzzy neural networks

Fuzzy neural networks (FNN) as opposed to neuro-fuzzy systems, whose main task is to process numerical relationships, can process both numerical (measurement based) information and perception based information. In spite of great importance of fuzzy feed-forward and recurrent neural networks for solving wide range of real-world problems, today there are no effective training algorithm for them. Currently there are two approaches for training of FNN. First approach is based on application of the level-sets of fuzzy numbers and the back-propagation (BP) algorithm. The second approach involves using evolutionary algorithms to minimize error function and determine the fuzzy connection weights and biases. The method based on the second approach was proposed by the authors and published in Part 1 of this paper [R.A. Aliev, B. Fazlollahi, R. Vahidov, Genetic algorithm-based learning of fuzzy neural networks. Part 1: feed-forward fuzzy neural networks, Fuzzy Sets and Systems 118 (2001) 351-358]. In contrast to the BP and other supervised learning algorithms, evolutionary algorithms do not require nor use information about differentials, and hence, they are most effective in case where the derivative is very difficult to obtain or even unavailable. However, the main deficiency of the existing FNN based on the feed-forward architecture is its adherence to static problems. In case of dynamic or temporal problems there is a need for recurrent fuzzy neural networks (RFNN). Designing efficient training algorithms for RFNN has recently become an active research direction. In this paper we propose an effective differential evolution optimization (DEO) based learning algorithm for RFNN with fuzzy inputs, fuzzy weights and biases, and fuzzy outputs. The effectiveness of the proposed method is illustrated through simulation of benchmark forecasting and identification problems and comparisons with the existing methods. The suggested approach has also been used for real applications in an oil refinery plant for petrol production forecasting.

Dynamic data mining technique for rules extraction in a process of battery charging

Battery charging controllers design and application is a growing industry direction. Fast and efficient charging of battery packs is a problem which is difficult and often expensive to solve using conventional techniques. The majority of existing works on intelligent charging systems are based on expert knowledge and heuristics. Not all features of the desired charging behavior can be attained by the hard-wired logic implemented by expert generated rules. Because the battery charging is a highly dynamic process and the chemical technology a battery uses varies significantly for different battery types, data mining technique can be of real importance for extracting the charging rules from the large databases, especially when the charging logic is to be continuously changed during the life of the battery dependent on the type and characteristics of the battery and utilization conditions. In this paper we use soft computing-based data mining technique for extraction of control rules for effective and fast battery charging process. The obtained rules were used for NiCd battery charging. The comparative performance evaluation was done among the existing charging control methods and the proposed system, which demonstrated a significant increase of performance (minimum charging time and minimum overheating) using the soft computing-based approach.

Linguistic time series forecasting using fuzzy recurrent neural network

It is known that one of the most spread forecasting methods is the time series analysis. A weakness of traditional crisp time series forecasting methods is that they process only measurement based numerical information and cannot deal with the perception-based historical data represented by linguistic values. Application of a new class of time series, a fuzzy time series whose values are linguistic values, can overcome the mentioned weakness of traditional forecasting methods. In this paper we propose a fuzzy recurrent neural network (FRNN) based time series forecasting method for solving forecasting problems in which the data can be presented as perceptions and described by fuzzy numbers. The FRNN allows effectively handle fuzzy time series to apply human expertise throughout the forecasting procedure and demonstrates more adequate forecasting results. Recurrent links in FRNN also allow for simplification of the overall network structure (size) and forecasting procedure. Genetic algorithm-based procedure is used for training the FRNN. The effectiveness of the proposed fuzzy time series forecasting method is tested on the benchmark examples.

Fuzzy time series prediction method based on fuzzy recurrent neural network

One of the frequently used forecasting methods is the time series analysis. Time series analysis is based on the idea that past data can be used to predict the future data. Past data may contain imprecise and incomplete information coming from rapidly changing environment. Also the decisions made by the experts are subjective and rest on their individual competence. Therefore, it is more appropriate for the data to be presented by fuzzy numbers instead of crisp numbers. A weakness of traditional crisp time series forecasting methods is that they process only measurement based numerical information and cannot deal with the perception-based historical data represented by fuzzy numbers. Application of a fuzzy time series whose values are linguistic values, can overcome the mentioned weakness of traditional forecasting methods. In this paper we propose a fuzzy recurrent neural network (FRNN) based fuzzy time series forecasting method using genetic algorithm. The effectiveness of the proposed fuzzy time series forecasting method is tested on benchmark examples.

Expected utility based decision making under Z -information and its application

Real-world decision relevant information is often partially reliable. The reasons are partial reliability of the source of information, misperceptions, psychological biases, incompetence, and so forth. Z-numbers based formalization of information (Z-information) represents a natural language (NL) based value of a variable of interest in line with the related NL based reliability. What is important is that Z-information not only is the most general representation of real-world imperfect information but also has the highest descriptive power from human perception point of view as compared to fuzzy number. In this study, we present an approach to decision making under Z-information based on direct computation over Z-numbers. This approach utilizes expected utility paradigm and is applied to a benchmark decision problem in the field of economics.

Recurrent Fuzzy Neural Network Based System for Battery Charging

Consumer demand for intelligent battery charges is increasing as portable electronic applications continue to grow. Fast charging of battery packs is a problem which is difficult, and often expensive, to solve using conventional techniques. Conventional techniques only perform a linear approximation of a nonlinear behavior of a battery packs. The battery charging is a nonlinear electrochemical dynamic process and there is no exact mathematical model of battery. Better techniques are needed when a higher degree of accuracy and minimum charging time are desired. In this paper we propose soft computing approach based on fuzzy recurrent neural networks (RFNN) training by genetic algorithms to control batteries charging process. This technique does not require mathematical model of battery packs, which are often difficult, if not impossible, to obtain. Nonlinear and uncertain dynamics of the battery pack is modeled by recurrent fuzzy neural network. On base of this FRNN model, the fuzzy control rules of the control system for battery charging is generated. Computational experiments show that the suggested approach gives least charging time and least T end -T start results according to the other intelligent battery charger works.

Decision making with imprecise probabilities and its application

Experimental evidence has repeatedly shown that the widely used principle of maximization of expected utility has serious shortcomings. Non-expected utility theory suggests more adequate models. However, in these models utility functions and probabilities are mainly considered as real-valued functions whereas in reality human preferences are imprecise being described in natural language (NL). Nowadays a methodology for dealing with second-order uncertainty, or uncertainty2 is not available, whereas, real-world uncertainties mainly fall into this category. In this paper we present an effective decision theory under uncertainty2 when the environment of fuzzy events and fuzzy states are characterized by imprecise probabilities. The proposed theory includes a non-expected fuzzy utility function represented by a fuzzy integral with fuzzy-number-valued fuzzy measure generated by imprecise probabilities. The suggested theory encompasses the classical utility based decision analysis, cumulative Prospect theory and Choquet expected utility on bipolar scales. We apply this methodology for solving a real-life business problem.

Soft Computing in Finance

The stock market is very attractive due to high expected profit. On the other hand it is very risky. This creates a need for intelligent stock-trading systems that are intended to help the investors make realistic prediction for taking optimal decisions. Conventional approaches address Regression and Time Series Analysis methods for stock market prediction [5,14]. These methods do not give expected results in situations when the data are influenced by subjective factors such as psychological, macro-economical, or political issues. Also we cannot ignore those factors at all, because technical indexes only are not capable of proper description of a complicated real-world environment. An effective stock trading system must use both qualitative and quantitative factors.

Soft Computing Applications in Operations Management

Problems in transportation logistics, such as routing, assignment, dispatching etc. are very complex combinatorial optimization problems and their solution is related with certain difficulties.