Tu Xuyan

Intelligent Autonomous Decentralized System (IADS)

In the paper, Intelligent Autonomous Decentralized Systems (IADS) is proposed. It is the development based on the combination of autonomous decentralized systems (ADS) and distributed artificial intelligence (DAI). The concept, model, architecture, performance, methodology and technology for design and implementation of IADS are discussed. The IADS based on Agent (Union-IADS, Market-IADS, Group-IADS) are presented. They have not only autonomous controllability and coordination, but also some humanoid DAI, such as activity, mobility, sensibility, reactivity and sociality.

SoftMan and agent

SoftMan is a new concept based on the production of distributed technique, intelligent robot and artificial life. SoftMan is the development of a mobile agent and is a virtual robot in the network environment. It inherited many research fruits of robot and agent. Therefore, it has the essential features of the intelligent robot and agent, and at the same time has life features. Comparing with the agent, SoftMan not only has more all-around humanized intelligence, humanized behavior and function, but also has the ability of environment identification, self-determination and free will. SoftMan has some digital life features which agent has none, such as activity, learning and evolution, inheritance, variation and emotion. The corresponding theory and technology fruits provided a good foundation and reference for studying SoftMan. At first, this paper introduces the technology of SoftMan and agent, and then discusses the relationship between SoftMan and agent from the viewpoint of structure model, function, implementation technique, and application example.

Research on Scheduling in Multi-Softman System with the Learning Mode Based on Genetic Algorithms

The concept, characteristics and models of Softman are discussed in this paper. An individual Softman overall model is given. Also the learning mode based on genetic algorithms which was used for the scheduling (decomposition of the task, distribution of sub-duties and multi-Softman parallel solution) in multi-Softman system is proposed. Genetic algorithms are mostly applied to the distribution of the task and in multi-Softman parallel solution

Approximate Solution Research of State Equation for Controlled Nonlinear Systems

Based on the solution of the linear state equation, by making of the Taylor expansion, the nonlinear state equation of controlled systems under ideal state is transformed to a set of ordinary differential equations with infinite series expression. Based on the solution of this set of linear state equation, the integral equation of the nonlinear state equation is also obtained by utilizing constant variation method. Finally, any order approximate solution of the nonlinear state equation is given by utilizing successive approximation method.

Research on the architecture of ERP system based on intelligent autonomous decentralized system

The defects of traditional ERP (enterprise resource planning) system are analyzed. The application feasibility of intelligent autonomous decentralized system to the structure of ERP system are discussed. The architecture of distributed intelligent ERP system is presented. It has some advantages such as intelligence, reengineering and flexibility. Moreover, the integration of material- flow, capital-flow and information-flow in ERP system is analyzed. The implementing model of ERP system based on Intelligent Autonomous Decentralized System is presented.

Autonomous decentralized organization model and coordination control of SoftMan Crowd

On the basis of the concept of Intelligent Autonomous Decentralized Systems (IADS) and SoftMan, the organization structural model and coordination control schedule of SMC (SoftMan Community) is put forward, SM.man management and control strategy is designed and the coordination control characteristics, forms and strategy of SMC are discussed. At the same time, the interacting control mechanism of local SMC and remote SMC is discussed.

Compression method based on training dataset of SVM

Abstract The method to compress the training dataset of Support Vector Machine (SVM) based on the character of the Support Vector Machine is proposed. First, the distance between the unit in two training datasets, and then the samples that keep away from hyper-plane are discarded in order to compress the training dataset. The time spent in training SVM with the training dataset compressed by the method is shortened obviously. The result of the experiment shows that the algorithm is effective.

Study on Artificial Emotion Model

Artificial emotion is a companion volume of artificial intelligence. Emotion is key to human action of decision-making. Emotion, intelligence and behavior have intimate relations. Emotion model is a vital question to the research of artificial emotion. This paper introduces generalized model of artificial emotion and emotional control model in softman control system. Erf (emotion restrain factor) and Eef (emotion excursion factor) are brought forward. Erf , Eef and V out are the three elements of emotional control in softman. Results of simulation prove that influence of Eef to emotional control is stronger than that of Erf. It matches the natural behaviors of human emotion

A bridge structural health data analysis model based on semi-supervised learning

Bridge structural health monitoring is a multi-parameter monitoring for guaranteeing safe construction and service of bridges. Focused on the features of the collected data by various front end sensors, that are reflecting bridge structural health state such as strain, vibration, distortion, cable tension etc., a bridge structural health data analysis model is established in this paper, based on semi-supervised learning which classifies diversified parameter data, and using classifier under various learning patterns, to conduct classification of two types of sample set respectively, on which analysis is done so as to diagnose the bridge structural damage degree and provide evidence and guidance to bridge maintenance and management decision taking.

Multi-softMan Coordination Model and Its Application in Metallurgical Construction Project Management

Soft-Man is put forward as a intelligent tool and technology which is humanoid and with higher intelligent level. Soft-Man can favorably cooperate within a Soft-Man group at various levels. Due to Soft-Man is the development of Object and Agent in software engineering, so that, the methodology and technology of software engineering will be developing from Object-Oriented (O/O), Agent-Oriented (A/O) to Soft-Man Oriented (S/O). Our research and development on Soft-Man based on two ways: Linux Kernel rebuilding and Java technology. The main results as follows: Concepts and Models of Soft-Man and Generalized Soft-Man; Personal Soft-Man Model, Architecture and moving; Soft-Man Group organization, management and cooperation; Soft-Man Society building, communication and coordination. The metallurgical construction project is characteristic of multi-target, huge investment and multi-contractor. It is an urgent problem to solve that to improve the Project Management level and efficiency, especially to take full advantage of Internet collecting information, exchanging and sharing information with each contractor. In this paper, intelligent management system for metallurgical construction project based on “Multi-SoftMan” is proposed. Its implementation platform and method are also proposed, and are considered to be feasible by Project Management representatives.