In cooperative multiagent systems, to learn the optimal policies of multiagents is very difficult. As the numbers of states and actions increase exponentially with the number of agents, their action policies become mo...In cooperative multiagent systems, to learn the optimal policies of multiagents is very difficult. As the numbers of states and actions increase exponentially with the number of agents, their action policies become more intractable. By learning these value functions, an agent can learn its optimal action policies for a task. If a task can be decomposed into several subtasks and the agents have learned the optimal value functions for each subtask, this knowledge can be helpful for the agents in learning the optimal action policies for the whole task when they are acting simultaneously. When merging the agents’ independently learned optimal value functions, a novel multiagent online reinforcement learning algorithm LU-Q is proposed. By applying a transformation to the individually learned value functions, the constraints on the optimal value functions of each subtask are loosened. In each learning iteration process in algorithm LU-Q, the agents’ joint action set in a state is processed. Some actions of that state are pruned from the available action set according to the defined multiagent value function in LU-Q. As the items of the available action set of each state are reduced gradually in the iteration process of LU-Q, the convergence of the value functions is accelerated. LU-Q’s effectiveness, soundness and convergence are analyzed, and the experimental results show that the learning performance of LU-Q is better than the performance of standard Q learning.展开更多
The necessi ty of product oriented integrated computer aided process planning (CAPP) is analy zed, and the system architecture and the main functions are described in detail. The key issues of the system, such as the ...The necessi ty of product oriented integrated computer aided process planning (CAPP) is analy zed, and the system architecture and the main functions are described in detail. The key issues of the system, such as the product oriented process planning me thod, the human machine cooperation in process planning, the generation of manu facture bill of material (MBOM) and the dynamic data exchanging technology betwe en CAPP and enterprise resource planning (ERP), are discussed. The CAPP system h as been applied at a CIMS environment corporation successfully.展开更多
How to manage and use models in DSS is a most important subject. Generally, it costs a lot of money and time to develop the model base management system in the development of DSS and most are simple in function or can...How to manage and use models in DSS is a most important subject. Generally, it costs a lot of money and time to develop the model base management system in the development of DSS and most are simple in function or cannot be used efficiently in practice. It is a very effective, applicable, and economical choice to make use of the interfaces of professional computer software to develop a model base management system. This paper presents the method of using MATLAB, a well-known statistics software, as the development platform of a model base management system. The main functional framework of a MATLAB-based model base management system is discussed. Finally, in this paper, its feasible application is illustrated in the field of construction projects.展开更多
With the development of information technology, DSS can be used to resolve the complex process of the feasible reasoning and scientific decision-making of projects. This paper offers 7 exploiting principles for the co...With the development of information technology, DSS can be used to resolve the complex process of the feasible reasoning and scientific decision-making of projects. This paper offers 7 exploiting principles for the computer support system on feasible reasoning and scientific decision-making of projects, that is, the principles of standardization, procedure, specification, agility, currency, practicability and development. On the basis of analysis on systematic procedure, the computer support system on feasible reasoning and scientific decision-making of projects is formed based on WEB, and its general structure, system function and the methods to be realized are introduced. The data composition of this system is analyzed following the principles of integrality, development, perspicuity and consistency. Also, the model-base management system is designed for the management of model storage and management of model operation.展开更多
Recognizing the drawbacks of stand-alone computer-aided tools in engineering, several hybrid systems are suggested with varying degree of success. In transforming the design concept to a finished product, in particula...Recognizing the drawbacks of stand-alone computer-aided tools in engineering, several hybrid systems are suggested with varying degree of success. In transforming the design concept to a finished product, in particular, smooth interfacing of the design data is crucial to reduce product cost and time to market. Having a product model that contains the complete product description and computer-aided tools that can understand each other are the primary requirements to achieve the interfacing goal. This article discusses the development methodology of hybrid engineering software systems with particular focus on application of soft computing tools such as genetic algorithms and neural networks. Forms of hybridization options are discussed and the applications are elaborated using two case studies. The forefront aims to develop hybrid systems that combine the strong side of each tool, such as, the learning, pattern recognition and classification power of neural networks with the powerful capacity of genetic algorithms in global search and optimization. While most optimization tasks need a certain form of model, there are many processes in the mechanical engineering field that are difficult to model using conventional modeling techniques. The proposed hybrid system solves such difficult-to-model processes and contributes to the effort of smooth interfacing design data to other downstream processes.展开更多
文摘In cooperative multiagent systems, to learn the optimal policies of multiagents is very difficult. As the numbers of states and actions increase exponentially with the number of agents, their action policies become more intractable. By learning these value functions, an agent can learn its optimal action policies for a task. If a task can be decomposed into several subtasks and the agents have learned the optimal value functions for each subtask, this knowledge can be helpful for the agents in learning the optimal action policies for the whole task when they are acting simultaneously. When merging the agents’ independently learned optimal value functions, a novel multiagent online reinforcement learning algorithm LU-Q is proposed. By applying a transformation to the individually learned value functions, the constraints on the optimal value functions of each subtask are loosened. In each learning iteration process in algorithm LU-Q, the agents’ joint action set in a state is processed. Some actions of that state are pruned from the available action set according to the defined multiagent value function in LU-Q. As the items of the available action set of each state are reduced gradually in the iteration process of LU-Q, the convergence of the value functions is accelerated. LU-Q’s effectiveness, soundness and convergence are analyzed, and the experimental results show that the learning performance of LU-Q is better than the performance of standard Q learning.
文摘The necessi ty of product oriented integrated computer aided process planning (CAPP) is analy zed, and the system architecture and the main functions are described in detail. The key issues of the system, such as the product oriented process planning me thod, the human machine cooperation in process planning, the generation of manu facture bill of material (MBOM) and the dynamic data exchanging technology betwe en CAPP and enterprise resource planning (ERP), are discussed. The CAPP system h as been applied at a CIMS environment corporation successfully.
文摘How to manage and use models in DSS is a most important subject. Generally, it costs a lot of money and time to develop the model base management system in the development of DSS and most are simple in function or cannot be used efficiently in practice. It is a very effective, applicable, and economical choice to make use of the interfaces of professional computer software to develop a model base management system. This paper presents the method of using MATLAB, a well-known statistics software, as the development platform of a model base management system. The main functional framework of a MATLAB-based model base management system is discussed. Finally, in this paper, its feasible application is illustrated in the field of construction projects.
文摘With the development of information technology, DSS can be used to resolve the complex process of the feasible reasoning and scientific decision-making of projects. This paper offers 7 exploiting principles for the computer support system on feasible reasoning and scientific decision-making of projects, that is, the principles of standardization, procedure, specification, agility, currency, practicability and development. On the basis of analysis on systematic procedure, the computer support system on feasible reasoning and scientific decision-making of projects is formed based on WEB, and its general structure, system function and the methods to be realized are introduced. The data composition of this system is analyzed following the principles of integrality, development, perspicuity and consistency. Also, the model-base management system is designed for the management of model storage and management of model operation.
文摘Recognizing the drawbacks of stand-alone computer-aided tools in engineering, several hybrid systems are suggested with varying degree of success. In transforming the design concept to a finished product, in particular, smooth interfacing of the design data is crucial to reduce product cost and time to market. Having a product model that contains the complete product description and computer-aided tools that can understand each other are the primary requirements to achieve the interfacing goal. This article discusses the development methodology of hybrid engineering software systems with particular focus on application of soft computing tools such as genetic algorithms and neural networks. Forms of hybridization options are discussed and the applications are elaborated using two case studies. The forefront aims to develop hybrid systems that combine the strong side of each tool, such as, the learning, pattern recognition and classification power of neural networks with the powerful capacity of genetic algorithms in global search and optimization. While most optimization tasks need a certain form of model, there are many processes in the mechanical engineering field that are difficult to model using conventional modeling techniques. The proposed hybrid system solves such difficult-to-model processes and contributes to the effort of smooth interfacing design data to other downstream processes.