In this paper, a necessary and sufficient condition is given for feasible solutions of a multilevel linear programming problem.Based on this condition, we design a type of simplex method efficiently to solve multilevel linear programes. We illustrate the algorithm via two numerical examples with three levels.

Agile manufacturing will be the strategy for manufacturing enterprise in 21st century. Agile manufacturing enterprise design means how to make it possible for an agile manufacturing enterprise to achieve the effective integration of its resources quickly in order to response the changes of market. The modern system science is the theoretical basis for agile manufacturing enterprise design.The traditional linear design model is no longer suitable to the design of complicate systems, while the sprial cycle model is able to meet the basic requirements of agile manufacturing enterprise design.

In the paper,the biological models is used to describe various advanced manufacturing systems. The models includes enterprise gene model,enterprise cell model,enterprise living-beings model,enterprises community model and industry bioecology systems model,all of which described the content,relationship and development direction of advanced manufacturing systems.

In this paper,illustrative examples are introduced to study the complexities,inclnding chaos caused by controllers.Two comments about simulation research and chaos in control systems are presented. Interesting phenomena that the control behaviour becomes irregular after seemingly convergent are proposed to show problems in simulation, with a further discussion to study the complex behaviours of systems with controllers.

In this paper, the multiperson hierarchical decisionmaking with bargaining followers is investigated in detail. The idea and method of incentive strategy designing are presented. As an example, the incentive strategy of LQ is derived, and the sufficient condition for existence of strategy is this case is given.

Strengthening the comprehensive demonstration, which was attention-getting in the Eighth Five-year Plan Period (1991-1995),is an important measure to upgrade the democratization and level of scientific management in decision making on the development for weapon systems and military equipment. This paper suggests that we must pay attention to strengthening the comprehensive evalution in the above-mentioned decision procedure. The necessity, goal, contents, methods,and operating mechanizm of the comprehensive evalution are revealed.

In this paper, we propose an exact method for solving a class of maximum entropy problems used in a decision making context.Comparing with the lagrange Multiplier method normally applied in the maximum entropy problems, this method is more simple and time-saving. It should be proved valuable for not only the decision analysis, but aslo the benchmark testing of the numerical nonlinear programming. Finally, an example about an investment decision problem is presented.

In this paper, a method of weights selected rationally is presented, provided that objective priority level is impossible to be determined by the previous experience. Based on the linearity weights, the energy function is used to simulate decision-making procedures.The evolution equation of the non inferior solutions and learning rule of weights is derived.It is proved from a analysis of the example that the method has a practical value to some extent.

The paper introduces the applications and problems of troditional computer software in organizational behavior briefly, puts forward the new method to solve the problems, that is to develop and build" the expert sytem building tool for organizational behavior". To use the tool,the domain experts can develop the expert systems for the domain conveniently by themselves. The paper introduces the basic principle and basic building methods of the tool, and how to use the tool to build new expert systems for the domain of organizational behavior.

A Regional Economic Forecasting Decision Support System(REFDSS) was developed in this paper, which used a method named"Tuning Method" to solve the forecasting problem in DSS of Regional Planning.The REFDSS includes four bases:Data Base, Method and Model Base, Knowledge Base, Graph Base.If the REFDSS is improved, it can be used in other fields.

From the view of macro economy, the authors analyse the mechanism that how gross national product(GNP),commodity prices, inflation, interest ate and exchang rate etc. influence stock market. At the same time they also analyse how stock market counteracts macro economy. From the analysis of relationship of macro economical coincident CI and Shanghai Stock index and Shenzhen Stock index,We can conclude that in our country there is high relation between stock market and macro economy and the trendcy of stock market precedes the one of macro economy.

This paper starts from that analysis of lossincurred in an enterprise and beneficial result is down,put research subject of Manufacturing Systems, meantime, analysis of malpractices and reform direction in an enterprise, and further discussed research aim, contents and method on new branch of learning.

This paper gives a simplex method on evaluating creditworthiness of every member belonging to an exchange after determining the relative importance of criteria with respect to the overall goal by AHP.

The optimum operation of a reservoir with hydropower station is a dynamic control problem which contains nonlinear objeotive function, linear and nonlinear constraints. Many methods have been developed for the optimization of reservoir management and operation.Most algorithms are based on some type of mathematical programming techniques.This paper presents a genetic algoritnm with floating point representation (FPGA) for optimizing operation of reservoir. The FPGA is suited for reservoir operation problem from both a mathematical and pratical point of view. Its main advantages lie in great robustness and less memory space. At the same time the combing FP representation with special operators can greatly improve GAS performance and such approach provides new means for overcoming dimensionlity curse of reservoir operation problem.

The noliner problem of rain flow process havebeen noticed by hydrologiers earily. Hydrologiers researched the hydrological noliner modeling and forecasting problem by means of experimential method hydrodynamics method, and system identification method in special. The authors of this paper submit an new method, combining the artificial nerve network model and the coder,code converter of information theory. Also, this paper discussed the approach of model used in flow short term forecast.

On the basis of analysing the shortages of the recent interconnet degree computation, this thesis puts forword the new computation model of the interconnet degree.The new computation model overcomes these shortages and possesses still characters of original model:uniquences, symmetry,comparable character. At the same time, the new model has new characters:normal property, non-coefficient effect.

Combined with the production and investment practice of coal electricity enterprises, this paper builds a dynamic input output model for corporate planning, including the planning of plant production capacities, the distribution of resources, and the dynamic balancing of production and investment.The application of model is illustrated by using it in a coal electricity enterprise for the optimization of investment structure and product mix, and the assessment of optimal corporate development strategies.

Applly the theories of both value engineering operations research, having optimal design indexes for either individual or comprehensive manufacturing technology.