Search

So far, many studies on Double-Deck elevator systems (DDES) have been done for exploring some more efficient algorithms to improve the system transportation capacity, especially in a heavy traffic mode. The main idea of these algorithms is to decrease the number of stops during a round trip by grouping the passengers with the same destination as much as possible. Unlike what happens in this mode, where all cages almost always keep moving, there is the case, where some cages become idle in a light traffic mode. Therefore, how to dispatch these idle cages, which is seldom considered in the heavy traffic mode, becomes important when developing the controller of DDES. In this paper, we propose a DDES controller using genetic network programming with idle cage assignment algorithm embedded for a light traffic mode.

Genetic network programming (GNP) is a newly developed evolutionary approach which can evolve itself and find the optimal solutions. A lot of research has been done and it has been demonstrated that GNP which has a directed graph structure can deal with dynamic environments very efficiently and effectively. It can be used in many areas such data mining, elevator supervising control systems, the strategy of buying and selling stocks in stock markets, forecasting the traffic volumes in road networks, etc. In order to improve GNPpsilas performance further, this paper proposes a method called GNP with Rule Chains. The aim of the proposed method is to balance exploitation and exploration, that is, to strengthen exploitation ability by using the exploited information extensively during the evolution process of GNP. The proposed method consists of 4 steps: rule extraction, rule selection, individual reconstruction and individual replacement. Tileworld was used as a simulation environment. The simulation results show some advantages of GNP with rule chains over conventional GNPs.

An algorithm capable of finding important time related association rules and its application to classification systems have been described in this paper. We firstly describe a method of class association rule mining using genetic network programming (GNP) with time series processing mechanism in order to find time related sequence rules. Secondly, the classification system is applied to estimate to which class the current traffic data belong based on extracted association rules. Using this kinds of classification mechanism, the traffic prediction could be done since the rules extracted are based on time sequences. And, we also present experimental results using the traffic prediction problem.

A new evolutionary method named ldquogenetic network programming with control nodes, GNPcnrdquo has been applied to determine the timing of buying or selling stocks. GNPcn represents its solutions as directed graph structures which has some useful features inherently. For example, GNPcn has an implicit memory function which memorizes the past action sequences of agents and GNPcn can re-use nodes repeatedly in the network flow, so very compact graph structures can be made. GNPcn can determine the strategy of buying and selling stocks of multi issues. And GNPcn can distribute the purchase capital to each stock based on the distribution ratio. The effectiveness of the proposed method is confirmed by simulations.

Classification rule mining is an active data mining research area. Most related studies have shown how binary valued datasets are handled. However, datasets in real-world applications, usually consist of fuzzy and quantitative values. As a result, the idea to combine the different approaches with fuzzy set theory has been applied more frequently in recent years. Fuzzy sets can help to overcome the so-called sharp boundary problem by allowing partial memberships to the different sets, not only 1 and 0. On the other hand, fuzzy sets theory has been shown to be a very useful tool because the mined rules are expressed in linguistic terms, which are more natural and understandable for human beings. This paper proposes the combination of fuzzy set theory and ldquogenetic network programmingrdquo (GNP) for discovering fuzzy classification rules from given quantitative data. GNP, as an extension of genetic algorithms (GA) and genetic programming (GP), is an evolutionary optimization technique that uses directed graph structures as genes instead of strings and trees; this feature contributes creating quite compact programs and implicitly memorizing past action sequences. At last, experimental results conducted on a real world database verify the performance of the proposed method.

In this paper, a method for traffic flow prediction has been proposed to obtain prediction rules from the past traffic data using genetic network programming(GNP). GNP is an evolutionary approach which can evolve itself and find the optimal solutions. It has been clarified that GNP works well especially in dynamic environments since GNP is consisted of graphic structures, creates quite compact programs and has an implicit memory function. In this paper, GNP is applied to create a traffic flow prediction model. And we proposed the spatial adjacency model for the prediction and two kinds of models for N-step prediction. Additionally, the adaptive penalty functions are adopted for the fitness function in order to alleviate the infeasible solutions containing loops in the training process. Furthermore, the sharing function is also used to avoid the premature convergence.

A lot of papers have studied how to select accurate association rules for the users, and most of them make use of efficient statistical measurements, such as support, confidence, chi-square, lift, collective strength, and so on. However, it is still an open question about how to select a more proper statistical measurement for a certain user to find association rules in a certain domain. Since different measurements consider different interesting aspects of the association rules, we propose a novel method, named RuleRank, combining these statistical methods, in order to provide an alternative with more flexibility and robustness for users. In order to capture the userpsilas interest more closely, we also integrate semantic similarity into RuleRank model. The simulation results show that RuleRank model could give satisfying results for the user.

Elevator group control systems are the control systems that systematically manage elevators in order to transport passengers efficiently. With the increasing need for high-performance transportation systems in buildings, multi-car elevators where two cars operate separately and independently in an elevator shaft are attracting attention as the next novel elevator system. Genetic network programming(GNP) can introduce various priori knowledge of the elevator systems in its node functions easily and execute an efficient rule-based group control that is optimized evolutionary. This paper discusses the development of controllers for multi-car elevator system (MCES) using GNP The effects for MCES are examined, and we compare the advantages and performances between MCES and double-deck elevator system (DDES).

A simple traffic system is proposed in this paper, in which the global optimal route considering the traffic volume is selected as the guidance for the vehicles. In the proposed system, Q value-based dynamic programming is adopted to calculate the optimal traveling time to each destination from every intersection of the road network. And Boltzmann optimal route method is used to generate several route candidates, the usefulness of which is evaluated in terms of the total traveling time considering traffic volume. Furthermore, dynamically changing traffic volumes of all the given origin-destination pairs of road networks are constantly provided in the road simulation we used and the traveling time of each section is continuously updated according to its dynamic traffic volume. In this paper, the analysis and comparison between greedy strategy and Bolztmann strategy with various ldquotemperaturesrdquo are carried out. The simulation result showed the effectiveness of the proposed Boltzmann optimal route method.

Many evolutionary computation methods applied to the financial field have been reported. A new evolutionary method named ldquogenetic network programmingrdquo (GNP) has been developed and applied to the stock market recently. The efficient Trading rules created by GNP has been confirmed in our previous research. In this paper a Multi-Brands optimization system based on Genetic Network Programming with control nodes is presented. This method makes use of the information from Technical Indices and Candlestick Chart. The proposed optimization system, consisting of technical analysis rules, are trained to generate trading advice. The experimental results on the Japanese stock market show that the proposed optimization system using GNP with control nodes method outperforms other traditional models in terms of both accuracy and efficiency.

This paper aims to solve multi-objective problems by adaptive random search with intensification and diversification combined with genetic algorithm (RasID-GA). Problems with multi-objectives are common in engineering, economics, computer science, and many others field of studies. It has been a challenge for the researchers to develop algorithms able to solve this kind of problem. RasID is an optimization algorithm, which is good at finding local optima, but its diversified search isnpsilat so efficient, for this reason, we combined RasID with genetic algorithms (GA), which is superior at finding global optima. In this paper, RasID-GA is used to find the Pareto- optimal solutions. RasID-GA is compared with the algorithm of NSGA-II using well known benchmarks.

One of the most important functions of car navigation systems is to assist drivers to travel to destinations. Minimum traveling distance or time is often considered to find the optimal route to the destination. However, when searching for the optimal route, it is essential to consider userspsila preference in the search. In this paper, we propose a new dynamic programming method to find the optimal route to the destination considering multiple criteria. The proposed method is an extension of our previous route search algorithm, Q method. In this study, we will define the Q values for the pair of road sections. A general index will be introduced, which takes multiple criteria into consideration. Q values are updated based on the local information available including other Q values at the adjacent road sections. When the traveling time changes, we use the already obtained Q values to find the new optional route. We applied the proposed method to a very simple artificial road network. Simulation results show that we can find better routes when considering multiple criteria compared to the optimization using the traveling time only.