Chen Ying-wu

Optimized road network emergency restoration scheduling for relief distribution

In order to minimize the length of time required for emergency relief distribution, the decision-maker always considers damaged road network restoration as a strategy. Focusing on the interrelationship between relief distribution and road network restoration, we provide a transparent and feasible solution for solving optimized road network emergency restoration scheduling for relief distribution. Combining the road network emergency restoration and relief distribution, a combinatorial optimization model is developed in this paper. Consequently, a newly developed genetic algorithm is designed to solve this problem. A case study is presented to highlight the significance of the proposed model as well as the efficacy of solution strategy.

A Spacecraft Launch Organizational Reliability Model Based on CSF

Organizational reliability, with regard to safety, needs to be assessed for the launch of spacecraft. In this paper, the compensatory and irreplaceable factor is defined while the organizational reliability of a spacecraft launch is researched by studying the decomposition and composition of factors. The factors of space and the characteristics of the organizational reliability of a spacecraft launch are analyzed. The composition method for the states of compensatory and irreplaceable factors is then established, and the overall organizational reliability model of spacecraft launch is constructed. This paper also uses the grey relational analysis technique to analyze and construct a model to assess the organizational maturity of important factors affecting organizational reliability. Finally, a case analysis demonstrates the application of the organizational reliability assessment model of a spacecraft launch, which validates the feasibility and effectiveness of the composition of states of factors. Copyright

Multi-Satellite Mission Planning for Environmental and Disaster Monitoring Satellite System

This article concerns the problem of managing China’s Environmental and Disaster Monitoring Satellite System (EDMSS) which is composed of multiple satellites with different image types and different operation constraints. In order to coordinate these different characteristic satellites to maximally satisfy the user’s requests, we propose a novel distributed mission planning architecture which holds both centralized and decentralized features. Its kernel is a centralized scheduling module which considers the universal resource constraints that all satellites included may share, while its external module is responsible for checking the special constraints related to each satellite and repairing the schedule correspondingly. The mission of EDMSS may include both spot target observations and polygon area target observations. To describe them in one model, we present a dynamic segmenting method to partition polygon areas based on the characteristics of satellite orbits, parameters of on-board sensors and the shapes of the polygons. Through coverage analysis and a profit calculation function, the two different kinds of observations can be scheduled under the same problem-solving architecture. We also propose a mission planning system based on the above techniques. Long range planning and daily scheduling are implemented in a rolling-horizon way.

A disaster warning model for a spacecraft launch based on information fusion and network inference

The launch of spacecraft should synthesise various incomplete and uncertain symptoms of extreme events to dynamically assess the possibility of disaster during the launch readiness process. However, there are still many technical problems in disaster assessment of spacecraft launch engineering. For example, the problem of how to effectively use the disaster symptoms is not addressed well yet and the fact that group decision-making can be easily lead to the minority opinion being ignored. Based on the extreme event symptom obtaining technique, a disaster warning model combining both virtual and real evidence is built through information fusion and network inference, using evidence theory and Bayesian theory. Using information fusion, the proposed model shows good performance with regard to complementary evidence, elimination of data redundancy and reduction of uncertainty. Using symptom information-based network inference, this model guarantees the identification of the correct minority. Finally, the effectiveness, correctness and feasibility of the model are verified by analysis of the ‘Challenger’ space shuttle explosion.

Approach of remotely sensed data processing task scheduling problem based on ant colony optimization

With the development of remote sensing technology, remote sensing data frequency-intensive has received and processed, the demand of remote sensing applications has kept an increasing growth. The management and planning for multi-source remote sensing data processing became very complicated with the evolution of remote sensed application requests. The way of effect management and scheduling can improve utility of processing resources and sufficiently exert abilities of remote sensing processing center. Based on the multi-objective optimization characteristic of the problem, this paper presents the mathematical model of the problem. An ant colony optimization algorithm is proposed for solving this problem. At last, experiments results show the effectiveness of our approach compared with the results of heuristic algorithm and simulated annealing algorithm.

ISRMDSS: An information security risk management oriented multi-agent system

Nowadays, information has become the fundamental element in our society, and the security of information asset has been paid more and more attentions. Recently, people try to resolve information security management problem using new technologies in artificial intelligence, such as agent and ontology. However, existing studies mainly focus on how to perform access control task using agents or formalize knowledge about information security using ontology, and problems like information security risk management has been overlooked. In this paper, we provide an information security risk management oriented multi-agent system (namely ISRMDSS) to fill this gap.

Analyses on schedule-cost coefficient correlation of spaceflight project based on historical statistics and its application

Abstract When analyze the uncertainty of the cost and the schedule of the spaceflight project, it is needed to know the value of the schedule-cost correlation coefficient. This paper deduces the schedule distribution, considering the effect of the cost, and proposes the estimation formula of the correlation coefficient between the In(schedule) and the cost. On the basis of the fact and Taylor expansion, the relation expression between the schedule-cost correlation coefficient and the In-schedule-cost correlation coefficient is put forward. By analyzing the value features of the estimation formula of the In-schedule-cost correlation coefficient, the general rules are proposed to ascertain the value of the schedule-cost correlation coefficient. An example is given to demonstrate how to approximately amend the schedule-cost correlation coefficient based on the historical statistics, which reveals the traditional assigned value is inaccurate. The universality of this estimation method is analyzed.

A neuron-inspired fuzzy relation model of dynamic systems and its learning algorithms

In view of fuzzy sets and their operations, three kinds of logic neurons, i.e., AND, OR and AND/OR neurons, are present in this paper. And those neurons can be classified into two types: weighted and relational. Using AND, OR and AND/OR neurons, a fuzzy relational model for dynamic system is provided as well as its learning algorithms. By a simple example, the soundness and the learning capability of the algorithms are verified.