Abstract:This paper studies the oligopoly with the existence of both network externality and economy of scale. With a simple but meaningful model, the substitution ratio among the network externality and economy of scale is calculated in influencing market price, quantity and firms＇ profit. It is discovered that （1） network externality has strong substitute to the economy of scale so that the Index of Comprehensive Effect of Scale （scale of production and that of consumption, ICES） is designed to reveal their influence; （2） ICES increases the consumers＇ surplus but may reduce firms＇ profit; （3） ICES can determine the full market coverage and it has an opposite effect to finns＇ profit when all potential consumers purchase the good

Abstract:Essential operation properties for supply chain dynamic scheduling in mass customization are described and defined. To relieve the dominant contradiction, train of thought for scheduling process is put forward on the basis of former work. Then this thinking is followed in the operation process of dynamic scheduling, and supply chain scheduling mechanism in mass customization is also analyzed. Optimization mechanism is introduced into the process while finding the scheduling solution, and ants optimization algorithm for supply chain dynamic scheduling is adopt- ed to devise and improve its particular algorithm. Characteristic of the new algorithm lies in that in computation process it can not only reflect unique requirements for the operation characteristic of this manufacture mode, but will also compromise the relief train of thought for the dominant contradiction. Feasibility of the new algorithm ultimately gets validated through a simulation

Abstract:Ant Colony Algorithm（ACA） is a novel simulated evolutionary algorithm which shows many promising properties and can solve Traveling Salesman Problem（TSP） efficiently. On the basis of analyzing the difference between VRP and TSP, an Adaptive Hybrid Ant Colony Algorithm（AHACA） is proposed to solve VRP, which is improved from basic ACA by improving the basic rules and integrating 2-opt local search method and C-W algorithm in order to decrease computing time and avoid stagnation behavior of basic ACA. Moreover, the problem of acquiring feasible solution is also discussed, and four resolutions such as Mass Ant, Feasibility Process of Approximate Solutions etc. are also introduced. Simulation results show that the AHACA is feasible and valid for VRP

Abstract:A distribution network for one warehouse and multiple retailers is considered in this paper. The amount of vehicle is viewed upon as an operational variable without upper bound, which is known as flexible vehicle. According to whether the truck is full, different strategies are adopted, viz. direct shipment and peddling shipment. Using the peddling shipment strategy, a model named WVRP with weight coefficient is built. Distinguished from many other models, the demand of each retailer is also important factor to be considered when routing schedule is made. The retailers with larger demands have priority to be visited earlier. Finally, a genetic algorithm named PB- GA using a special partition method is proposed to solve this model, detailed comparison is made with usual heuristic for vehicle routing problem.[

Abstract:Traditional process capability analysis takes assumption that there is only a variation source in production/process quality characteristics. This paper presents a method for estimating process capability indices, which roots in nested mode of experiment design, and uses variance components to estimate total variance as well as each variance component. In addition, the confidence interval of Cp is derived based on MVA. A case is studied, and it is shown that MVA is good at analyzing the process variation patterns, so the method can remark the process scientifically and rationally.[

Abstract:Optimal investment time of lumpy capacity espansion investment under bistochastic uncertainties of price and capacity utilization rate was investigated in this paper. Applying theory of real options, it developed an investment decision-making model aimed to maximize investment options based on expected first hitting time. A simulation optimization algorithm designed by Riskoptimizer software package was used to solve the model. Numerical examples were carried out in the end which indicated that the biuncertainties＇ influences on optimal capacity expansion decisions

Abstract:Cramer pointed out shortcoming of plain Logistic default model and put forward bounded Logistic model. This paper does some further research about bounded Logistic default model. We first demonstrate why bounded Logistic default model is super to plain Logistic model theoretically through sampling distribution analysis. Then we give empirical evidences based on China companies＇ data. We not only find evidences about Cramer＇ s problem, but also find that bounded Logistic model can solve the Cramer＇ s problem, not sensitive to critical value and have higher prediction efficiency.

Abstract:This paper considers the pricing problem of Shanghai Copper futures with frictions related to costs of transaction, storage, payments and margin. First we propose the pricing formula for futures asset： the difference of prices generated by these formula with the prices of traditional non-arbitrage pricing method under the no-frictions assumption is that the price here is a range, rather than a point. Then we test the formula with 78 contract data of Shanghai Copper futures from 9601 to 0206. The results show that the frequency for the real prices of Shanghai Copper futures falls on pricing range are 7.83 %. It is also shown that as time passes by, however, the frequencies are becoming more and more high. This indicates that Shanghai Copper futures markets are becoming more and more regulated, and the trading efficiencies of the markets are increasing

Abstract:From March 3rd 2003, a new tick size policy of closed-fund has been taken into effect at Shanghai and Shenzhen Stock Exchanges, which changed from ＂1 fen＂ to ＂0.1 fen＂. The new policy is aimed to improve the closed-fund market liquidity and ＂splint arbitrage mode＂. By comparing the difference of three market quality measurement indicators including bid-ask spread, depth and volume before and after the tick size changed, we found that to some extent the ＂0.1 fen＂ tick size improved the closed-fund market quality such as liquidity and trading activity

Abstract:An important assumption lies behind the efficient-market hypothesis： the market is so efficient that every new information is spread to the whole market immediately, and the latter takes rational and ＂correct＂ reaction to the new information at once. The prices of the securities reach new balance at the time the news is released. However, this doesn＇t exist in reality because it takes time for the news to travel and for the market to react. In order to make improvements to the efficient-market hypothesis, and facilitate the quantitative measurement of the efficiency of the securities market, we put forward the Spring Oscillator Theory, which advances the test of the random-walk model within the frame of efficient-market hypothesis from the 1st order to the 2na order. This model expresses the weak efficient market and the sub-strong efficient market using the unified fluctuation equation. The Theory can better explain the ＂over-reaction＂, ＂under-reaction＂ and the booming and collapse of the stock market with no news

Abstract:Swarm Intelligence is the global intelligent behavior emerged from the interaction of groups of simple agents. From the view of complex systems, this paper comprehensively discusses the system structure, operation mechanism, modeling tools, algorithmic model and applications of swarm intelligence according to its fundamental principles. Firstly, around the system structure of swarm intelligence represented especially by ant colony and bird flock, the agents＇ attributes, their behavior rules and their interaction modes are analyzed, and then the feedback mechanisms and learning mechanisms implied in swarm intelligence are induced and revealed. Based on the intro- ductions to the commonly-used modeling tools such as Genetic Algorithm （GA）, Artificial Neural Network （ANN）, Cellular Automata （CA）, and Agent-Based Modeling （ABM）, etc. , four kinds of typical models and algorithms of swarm intelligence, viz. , ant colony foraging, ant clustering, labor division in ant colony, and bird flock foraging, are discussed in detail, aiming to conclude the general rules to model and simulate thecomplex systems based on swarm intelligence. Finally, the applications of swarm intelligence to engineering optimization, production management, robotics, data analysis and pattern recognition are introduced, and some perspectives on the development of swarm intelligence are made as the concluding remarks of this paper.