This paper proposes a mean-CVaR portfolio optimization framework for both conservative and aggressive investment strategies based on long-term risk prediction, which is different from financial assets such as stocks, bonds portfolio optimization framework based on the short-term risk prediction, so as to mitigate concentration risk due to sharp fluctuations of price of single inventory in supply chain finance. The long-term risk prediction based on Monte Carlo simulation of the inventory portfolio is proposed, and it is more practical than square root rule, which overcomes the shortcoming of the square root rule which heavily depends on the independent normal distribution. In methodology, AR(1)-EGARCH(1,1)-EVT model is set up to better depict the characteristics of the autocorrelation, heteroskedasticity, leptokurtosis and fat-tails of the marginal distribution, furthermore, the multivariate t-Copula function is introduced to model the dependency structure of individual pledged inventory. The empirical results show that, the mean-CVaR optimization framework outperforms the improved mean-variance from the perspective of long-term risk prediction, which are robust to the choice of risk window, confidence level, simulation times and sample size. In summary, this paper provides a new framework for managing the risk of portfolio in inventory financing practice for banks constrained by risk limitation.

Both stock selection and optimal portfolio choice play crucial roles in setting portfolio. This paper proposed a new method for stock selection based on constructing a Sharpe index using intra-day high frequency data. Several portfolio strategies are considered too. An out-of-sample empirical analysis with respect to a sample of A shares listed on Shanghai stock exchange shows that the new stock selection method often offers higher risk-adjusted returns and lower levels of risk. Moreover, combined with optimal portfolio strategy, the new method can get considerable excess returns.

Taking the ten most important stock markets indexes as examples, this paper calculated the predicted VaR value of the portfolio with the four vine copula models, R-vine, D-vine, C-vine and R-vine all t, based on the rolling technique of monte carlo simulation. Furthermore, using rigorous Backtesting method, we compared the performance of forecasting VaR of the four vine copula models. The empirical results show that, the R-vine is the best model in measuring and predicting of the portfolio of the VaR no matter under the condition of equal weight or mean-CVaR. Under high quantile condition, the performance is better. Moreover, the overall prediction accuracy of D-vine is higher than that of C-vine and R-vine all t, however, the performance of R-vine all t is relatively worse when there is only t copula between the nodes.

This paper takes advantage of the Baidu News search engine to extract data of internet news about public companies. We test the determination of the cross-section of expected stock returns in Chinese A-share market from 2003 to 2010, and the empirical results turn out to be consistent with Fang etc. To unveil the intrinsic mechanism of "media coverage effect", we decompose media attentions into expected portion and unexpected portion. The result suggests that the "media coverage effect" of stock return mainly contributed by the unanticipated part, yet the pricing effect of the unexpected attention can be explained by "heterogeneity investors opinions based on short-sales constraints" by Miller, and, likewise, the "speculative attention" hypothesis in this paper.

Considering systematic jumps and heterogeneity jumps as tail events, we discuss the tail characteristics of distribution of stock return from the perspective of the extreme value theory. We use time of day (TOD) method to eliminate intraday effect of high-frequency data, apply index-stock method to decompose systematic jumps and heterogeneity jumps, and the peak over threshold (POT) method to estimate the left tail and right tail parameters. Empirical studies show that the intraday effect of A-share market possesses apparent "L" type feature. Each stock exists significant systematic and heterogeneity jumps. And the tails of two jump types are obvious thick. The times and contributions of right tail jumps are larger than left in all stocks. This suggests that the frequent appearance of jumps and jump tail characteristics are an important reason for non-normal distribution of stock return.

Based on the online learning character of weak aggregating algorithm (WAA), this paper explores its application to online portfolio selection, and considers the situation of making decision according to finite expert advices. The WAA is first applied to the expert strategies that invest only on one stock; then the single aggregating strategy (SAS) for online portfolio selection is obtained and the competitive performance of this strategy is analyzed, which shows SAS can pursue the best stock. In real investment decision-making, investors may choose several stocks to construct portfolios to invest, the WAA is further applied to the expert strategies that invest on different numbers of stocks; the mixture aggregating strategy (MAS) for online portfolio selection is then obtained; the conclusion that the cumulative gain MAS achieved is as large as that achieved by the best expert advice is proved. Numerical examples of long-period portfolios are provided to illustrate that SAS can achieve gain as well as the best stock; MAS can achieve gain as well as the best constant rebalanced portfolio (BCRP) strategy, and can obtain more when compared with universal portfolio (UP) strategy, which shows great competitive performance.

This paper studies a long-term of crude oil markets and trend of dynamic weekly price during 1986-2012. Based on the discussion of multiple equilibria theory, the paper tries to explain the long cycle phenomenon of crude oil price fluctuations by taking close look at the evolution of the long-term market supply and demand. Resorting to structural breaks test and cumulants test, a jump diffusion model is investigated. Historical data analysis shows that the crude oil prices were characterized by high volatility, high intensity jumps, and upward drifts. Furthermore, the models forecast that even given the current sharp rise in the investment in the oil development, crude oil price will still remain at a high frequency jump and stay in jump for the next couple of years.

In this paper, we examine the characteristics of private benefits behavior of control by different modes of the separation of control rights from cash flow rights under the framework of ultimate owner agency. The results show that: 1) the level of private benefits of control has a positive correlation with the separation of the two rights, a negative correlation with cash flow rights of the controlling shareholder and a negative correlation with the degree of legal protection of shareholders' rights; 2) in a dual-class equity structure, the level of private benefits of control of the ultimate owner increases with multiples of voting rights of the superior class by diminishing marginal speed and the most common arrangement is a 10:1 structure in which the superior class has ten votes per share and the inferior class has one vote per share; in a pyramid equity structure, the level of private benefits of control of ultimate owner increases with tiers of the pyramid by diminishing marginal speed and the ultimate owner usually controls the operating company by a three-tier pyramid control chain; in a cross-ownership structures, the level of private benefits of control of ultimate controller increases with the number of companies within the group and the ultimate controller tends to create a huge business empire; 3) the numerical results show that the levels of the private benefits of control of the ultimate controller of China's listed companies are between 6% and 8%.

This paper studies how a manufacturer motivates its supplier to effectively expand their capacity investment under demand uncertainty, targeting at the conflict of capacity investment between the manufacturer and the supplier. In our model, the manufacturer determines contract parameters, whereas the supplier decides the magnitude of capacity investment. We consider the centralized decision as a benchmark and the wholesale price contract as a basis for comparison. Then we design two incentive mechanisms, i.e., revenue sharing and penalty, from the perspective of revenue and capacity, respectively. The study indicates that the penalty contract brings the optimal level of capacity investment, and that the revenue sharing contract fails to coordinate the supply chain. We further analyze the reason why the revenue sharing contract does not coordinate the supply chain. Finally, we design a modified contract of revenue sharing, and demonstrate that the modified contract could motivate suppliers to build a higher capacity and achieve coordination.

The gap between attitude and behavior is obvious in the green purchase behavior of rational consumers. Took a non-food environmental friendly product as an example, based on the abstracted properties and behavior rules of consumers' green purchase behavior, to construct interactive heterogeneous artificial consumers using computational experiments, and simulate the real behavior of consumers' green purchase. By the analysis of macroscopic individual and the "emergence" phenomenon, to explore the evolution rules for consumers' green purchase. The results showed, green purchase forerunners played an important role in improving green market share by doing positive propaganda works. Government subsidies for green products can significantly improve the green products market share in a short term, but the long-term effect was dependent on the dissemination of information. Green consumption integral system can maintain a high and long-term market share for green products.

It is usually assumed that different lines of business are independent, but the fact is that they are dependent to some extent in multi-lines of business. Their dependence may be captured by Vine Copula functions. Vine Copula is a powerful tool to solve multiple dependence. Under the assumption that the incremental paid claims of every line of business follows gamma distribution, inverse-Gaussian distribution and log-normal distribution, respectively, the corresponding Vine Copula regression models are established. The model is applied to a real data set of auto insurance and the result shows that the Vine Copula-based regression model is superior to independent regression models in claims reserving.

Based on the political economics framework, this paper introduced the heterogeneity of local governments into the promotion tournament which leading by central government, improved the traditional promotion game model, and comparatively analyzed the promotion tournament under the conditions of homogeneity and heterogeneity of competitors, this paper had these conclusions: first, remuneration gap between winners and losers was a strong incentive factor, competitors heterogeneity will affect the level of their efforts; second, the heterogeneity of competitors made the output fluctuation no longer always be a bad factor for disadvantage competitors, because the output fluctuation brought disadvantages competitors over advantage competitors opportunity space; third, this paper explained the behavior differences of local governments in the phenomenon of "west strong east weak" on China's economic growth and provided some beneficial implications for governments management reform.

This paper is to investigate social loafing in knowledge-based virtual teams. Taking social capital theory and motivation theory as research framework, the paper presented a model to explain the influence mechanism of social loafing. Both individual intrinsic motivations and circumstantial conditions were examined. Quasi-experiment and survey method among 11 teams comprised of 87 students were employed to test the model. Results demonstrate that individual's social capital significantly decreases social loafing. Individual's reputation and perceived social presence were found mediating the relationships between social capital and social loafing. Meanwhile, individual's goal striving has a significant influence on social loafing. Suggestions on team management, design of communication technologies and future research are discussed at the end.

Modeling of public opinion is no longer confined to the nereaset neighbourship model (‘person-person’effect), we introduced the influence of next nearest neighbours (‘person-person-person’effect). We proposed an opinion evolution model of social cascading impact, characterized the effect of neighbours' neighbours and investigated the opinion dynamics on tunable clustering network. By adjusting the triad formation parameter which was used to change clustering coefficient of network, we applied asynchronous updating mechanism to observe the clustering coefficient influence on opinion formation. Simulation results show that: 1) compared with traditional nearest neighbour impact model, the social cascading impact model has stronger social reinforcement and as a result, is easier to reach a dominant consensus. 2) when p₊≠ p_- in initial state, a large clustering coefficient favors development of a consensus; when p₊ = p_- in the initial state, consensus results present uncertainty.

Focused on spliced networks, new representations were designed, which equivalently represent all spliced relations (time lags) between activities by using classic critical path method (CPM) networks with activity-on-arc representations, and time parameters were computed by using algorithms of the CPM networks. The new representations make the spliced networks be intuitive and convenient like the CPM networks with activity-on-arc representations, and furthermore greatly simplify calculation procedures which provided by construction industry. In additional, by applying the new representations, strange phenomenon were discovered in the spliced networks, for example, the project duration will be prolonged instead of being shortened if shortening durations of some critical activities; and floats of some non-critical activities will be unchanged instead of changing no matter whether prolonging or shortening their durations; and so on. The strange phenomenon cause new challenges to solve project scheduling problems etc., and open up new field of the spliced networks in theory and application.

Process structure uncertainties in complex products R&D projects are analyzed. The R&D process is simulated and modeled with design structure matrix (DSM). DSM scheduling policies are used in the DSM simulation. We develop a multi-objective optimization model with the objective of minimizing both project duration and cost. We propose a hybrid algorithm based on DSM simulation and nondominated sorting genetic algorithm II to generate the Pareto-optimal solutions for our multi-objective optimization problem. The proposed hybrid algorithm is applied to a real-world R&D project process optimization problem. The results show that our hybrid algorithm is robust and valid. Moreover, a simulation is used to evaluate and analyze the resulting optimal project process structure.

The one-dimensional cell transmission model cannot depict different queue formations and dissipations at a multiphase signal controlled intersection. To separately depict vehicle queues at a multiphase-signalized intersection at microscopic level, a set of simple two-dimensional cell transmission logics on a link and at a signalized intersection is put forward. To properly represent the traffic dynamics at a multiphase signal controlled intersection, a simple two-dimensional cell transmission model is derived from the proposed logics based on Daganzo's cell transmission model. Finally, a series of models are put forward at cell level to dynamically estimate performance indexes of a traffic signal. Case studies show that the simple two-dimensional cell representation is better than the one-dimensional cell representation in that the estimation accuracy of the simple two-dimensional cell transmission model is higher than that of the one-dimensional cell transmission model, and that the simple two-dimensional cell transmission model can accurately represent traffic dynamics under saturated conditions. The proposed model can be used to evaluate the coordinated traffic signals.

Taking the traffic situation in the area of new small spacing interchange as a research object, it used traffic conflict technique and VISSIM traffic flow simulation technique to establish traffic conflict simulation and to obtain the small spacing interchange collision rate and spacing relational model; conflict ratio was taken as the safety assessment index in the small spacing interchange by means of gray cluster theory. According to the level of security in the small spacing interchange area, the paper divides into four types by the diffident spacing: unsafe (D-level), criticality safety (C-level), safer (B-level), security (A-level), and the two interchange spacing of less than 700m should be put forward corresponding measures to improve the traffic safety. Application results show that: Interchange of different spacing is not the same level of security, the use of improved measures can improve the safety situation of small spacing interchange area. At the same time, it provides effective safety evaluation of interchange for the new small spacing area.

Airline fleet composition formed by traditional fleet planning methods could not well adapt to the fluctuations of air demand. This paper divided time range into several time intervals for each route in the whole network by air demand fluctuation patterns. The frequencies of aircraft types flying on routes within time intervals were regarded as decision variables; Several factors including airworthiness limitations of aircraft types flying on routes, air crew available flight time and the least aircraft number of selected aircraft types in one fleet were treated as constraints; A model for optimized match between routes and aircraft types within time intervals was constructed, which considered the total operating profits of allocating appropriate aircraft types onto each route as objective function. Finally, Lagrange relaxation algorithm was used to solve airline fleet planning problem. A certain airline case including 19 routes, 299 flights and 6 candidate aircraft types indicates that this proposed approach can reflect the distributions of aircraft types on different routes. Furthermore, the formed fleet composition is more adaptability to the change of airlines' production environment compared with the previous approaches. So the model is feasible.

In order to be more reasonable of site selection decision-making of mechanical parking system in urban, give full play to its to alleviate traffic after the completion of the action. A method of site selection decision is proposed. Firstly, by using GIS (geographic information system) analysis method, the attribute information database sets of planning parking system are established. Secondly, applying fuzzy rough sets mutual information attribute reduction algorithm to be mined the geography knowledge between parking system's location and a variety of geography factor using attribute reduction algorithms. Then, with decision tree and decision rules generated by geographic information analysis and deductive reasoning combination of inductive learning algorithm, location decision-making is evaluated and analyzed. Finally, a case study of Lanzhou is presented, and the results show that this method not only can do well in site selection decision of urban and so on, but also provide a reference for intelligence transportation system.

In this paper, we propose a new algorithm multi-label based ant colony system (MLACS) algorithm to address the vehicle routing problem with heterogeneous fleet, simultaneous pickup and delivery (VRPHSPD) problem which meet often in the real word. Leveraging the object-oriented principle, we build multi-attribute labels for various customers (demands), vehicles and routes. And then, we initialize the solution through nearest neighbor heuristic approach and minimize the number of vehicles required and total travel length by MLACS. To evaluate the efficiency and effectiveness of MLACS, we compare it with other optimization algorithms using benchmark problems and practical problem. Our results show that the MLACS algorithm has significant advantages in achieving the objectives as well as computing time. In addition, a numerical simulation using the actual data shows that MLACS is also good at solving the real world problem.

With identity authentication becoming an essential mechanism to ensure robust system security in information systems, RSA-based authentication protocols have been studied intensively for their great practicality. This paper points out that a recent RSA-based remote user two-factor authentication protocol proposed by Xie et al. cannot achieve the claimed security and reports its following flaws: (1) It is vulnerable to replay attack and key compromise impersonation attack; (2) It suffers from the problem of user privacy violation and poor repairability. As our main contribution, an improved scheme is put forward and formally proved secure under the RSA assumption in the random oracle model. As compared with other related schemes, our scheme is the first one that can achieve provable security while keeping the merit of high performance. Consequently, our scheme is more well-suited for mobile application scenarios where resource is severely constrained and security is particularly concerned.

According to the influencing factors analysis of roads' emergency earthquake evacuating in key area, and study on basic ideas and principles of system dynamics, the dynamic feedback of earthquake evacuation was analyzed from three aspects that disaster environment, road space, and personnel characteristics, then the road evacuation causal model and flow model based on system dynamics were established. Complied with the standard for urban planning on earthquake resistance and hazardous prevention for emergency shelter, evacuation roads and evacuation time, dynamic simulation on the road evacuation capabilities of crowded emergency evacuation area during the earthquake was conducted, related recommendations were given for key personnel-intensive area planning from the simulation results, to provide a reference for urban earthquake relief work.

With analysis of the features of decision-making information fusion and actual emergency decision making, this paper presents a decision-making information fusion method based on distance entropy according to the demand for objectivity and scientificity. Firstly, the fusion set is acquired and stored based on knowledge unit model. Secondly, the information units are formed through instantiating knowledge units, which can provide information support for the fusion. Afterwards, the concept of distance entropy is proposed with an element of distance added into the traditional entropy weight method. Thus, we can obtain both local and global fusion weight, finally to gain the global fusion results using the way of linear weighting. Furthermore, the application value and effectiveness of this method are illustrated by empirical analysis. Experimental analysis show that the proposed method can solve problems of the hugeness of knowledge base and improve the objectivity and scientificity of fusion results in decision-making information fusion.

Quantum-inspired evolutionary algorithm has premature and slow convergence shortcomings on solving numerical optimization problems. To overcome these shortcomings, a novel quantum-inspired evolutionary algorithm based on culture & knowledge is proposed by introducing the cultural algorithm. This algorithm contains two evolutionary layers: quantum evolutionary layer and knowledge evolutionary layer. Since the introduction of cultural algorithm, this algorithm can achieve fine balance between exploration and exploitation as well as can escape from local optimum. Because of the new framework and quantum observation, the proposed algorithm not only retains the advantages of quantum coding, but also effectively solves numerical optimization problems. The experimental results show that the algorithm has better performance than the quantum-inspired evolutionary algorithms. The proposed algorithm performs better than other related algorithms in terms of speed and accuracy.

The multi-skill call center staffing problem is studied systematically in this paper. Erlang-A model, in which, the abandoning customers can be taken into consideration, was applied to estimate the need of operators. Then, a dynamic set covering algorithm was proposed to solve the multi-skill routing problem. Thereafter, a heuristic algorithm was developed to solve the large-scale call center staffing problem. Finally, the data of a real call center was utilized to verify the efficiency and effectiveness of the algorithm. In addition, a series of computational experiments were conducted to compare the service level of different personnel-grouping schemes and different working-hour-split schemes. The calculating results show that the smaller personnel-group and shorter working hours can improve the staffing quality. And compared with the personnel-grouping schemes, working-hour-split schemes can be more effective on the staffing quality improvement.

Evacuation in an emergency is a complex behavior. When evacuation individuals are intelligent agents, an intelligent decision-making model is necessary to describe such a behavior. This paper elaborates how a modular model based on cellular automation is constructed with the advantages of information expression of cellular automation and behavioral expression of robots. The description of evacuation is realized through the assembly of such modules as evacuation time model, direction selection model and passage attraction model. When the model is used for describing the evacuation process of a hypermarket, the analog simulation and the data analysis show that it is effective in the descriptions of individual and group evacuations.

Formal concept analysis is widely used in the field of software engineering and machine learning. The construction of formal concept is a prerequisite of the application. In order to make the construction of formal concept intuitive and simple, this paper proposes a formal concept construction algorithm based on attribute topology. Using the mutex and including relationship among the attributes, the formal concept is constructed by the definition of "road". Some properties of "road" in attribute topology are founded by studying it, and used to optimize the algorithm, which improves the operational efficiency. Experimental result validates the feasibility of the proposed method.

This research investigates a k-out-of-n system in which the operational, standby and repair time of units follow continuous Phase-type distribution, instead of exponential distribution or other typical distributions. The proposed model provides an analytic reliability model that is more suitable to characterize the real situation. The stationary distribution is built by using matrix analytic methods; and several performance measures of interest, such as the system stationary availability, operational time, mean time between failures (MTBF) and failure arrival rate, are obtained. Finally, the validity and applicability of the model are implemented by numerical applications; and the influences of k on the system reliability laws are demonstrated when the value of n is fixed.

According to the difficulty of quantifying the uncertain index in the combat effectiveness evaluation of submarine-to-air missile weapon system, the grey cloud model is applied to improve the classic method of determining whitenization weight, which solves the transition problem between qualitative and quantitative. Moreover, three scale methods cited to construct the judgement matrix overcome the disadvantage that the classic analytic hierarchy process (AHP) is not objective and needs consistency examination. On this basis, a new combat effectiveness evaluation model which combined grey cloud model and improved analytic hierarchy process is built up. This model increases the credibility of effectiveness evaluation and provides a new valid approach to combat effectiveness evaluation of submarine-to-air missile weapon system.

Based on comparison between classical decision-making issues and crisis decision-making issues, specified constrain conditions of crisis decision-making are analyzed, which behaves more like a complex system. In this paper, a decision-making method for complex crisis is developed, which consists of three principles, a cross-simulation model, and a decision-making model based on cross-simulation. The three principles include the dynamic anchoring principle, the effective non-optimal principle, and the bottom line principle. The three principles aim at three major difficulties of crisis decision-making, which are no-fixed objective, lack of evaluation method, and hard to adjust. The cross-simulation model contributes to accuracy and reliability of simulation with integration of data and models. The decision-making model based on cross-simulation provides effective decision support for complex crisis. The method developed in this paper may be applied to more complex issues not limited to crisis.

The parallel emergency management has become an effective method to emergency management and the computation experiment is the key capacity of this method. Firstly the parallel emergency management's basic idea and the dynamic model of the KD-ACP platform were introduced. Then the parallel emergency management oriented computation experiment frame was designed. Based on the generic principles of the experiment frame and the characteristic of parallel emergency management, the process model of computation experiment was proposed. Finally the experiments of H1N1 transmitting in Beijing city based on the experiment framework in KD-ACP were executed and four levels' emergency response measures were designed, and the experimental results were contrasted. According to the application, the proposed frame can support the parallel emergency management very well.

Using the data in the influenza A H1N1 positive rate of sampling from August 3, 2009 to November 14, Beijing, we estimated the precision of absolute error and relative error. During the epidemic, absolute error of the H1N1 positive rate was less than 0.1 under the confidence level of 1-α=0.95, this error was at a controllable level. The relative error of the H1N1 positive rate was large at the early phase of influenza A H1N1 transmission, > 0.5 before 37 weeks. With the development of the epidemic, the relative error decreased gradually, which means that the result tends to be a reliable estimate. The relative error was less than 0.2 after 42 weeks. Finally, this paper provided a sample size planning based on precision control for sampling test of positive cases. We argued the marginal changes in sample size with a series of conditions. Optimal sample sizes were estimated under different combinations of positive rate and absolute error.

Existing evacuation models consider less the influence of heterogeneous crowds motion on evacuation efficiency. In this paper, a new mesoscopic evacuation model based on multi-agent and entropy is proposed. The mesoscopic model includes the upper macro multi-objective evacuation path optimization model and lower microscopic pedestrian evacuation model. The concept of evacuation entropy is introduced to construct the evacuation entropy map which reflects the degree of crowds movement disorder and can affect individual behavior. The upper model is a multi-objective optimization model based on quickest flow model, which calculates the global optimal evacuation path for individual. The lower model includes entropy-based grouping behavior model and leading behavior model. The simulation experimental results show that leaders in crowds have significant impact on evacuation efficiency. When the number of leaders increase in crowds, the number of crowded behavior, obstacle avoidance behavior and crowd panic level are reduced. The proposed model can make the evacuation process more orderly and reduce evacuation entropy.

Emergency management is an issue of complex systems which lack effective solutions. Due to the challenges of the low frequency, unclear evolving models and high uncertainty during an emergency, traditional simulation-based models and data-based models, which are widely used in normal states, are not very applicable in emergencies. In this paper, a simulation method is proposed to provide simulation systems with the adaptability of simulating normal and emergency states, through the combination of simulation models and big data technologies. Using a scenario of the large-scale traffic evacuation as an example, the simulation method improves the accuracy and the diversity of prediction, and supports decisions of the emergency management for traffic evacuations.

The research on emergency management needs novel modeling paradigm to solve the modeling of emergent events. The emergency management oriented meta-modeling approach describes the basic elements of artificial society by meta-models. With the help of these meta-models, the emergency management models are built first, and the simulation models are generated by model transformation in order to support computational experiments. We choose a public health event as the case to study how to build the meta-model of the artificial society. Accordingly, emergent event model and emergency management model are also built by the meta-modeling approach in this paper. The code framework of domain models is generated by the simulation model code generation technique. The construction of models in this case illustrates the process of emergency management oriented meta-modeling approach. The case testifies that the modeling approach of artificial society is self contained, and it plays an important role in the integration of emergency management research project.

The departments involved in emergency decision-making are required to perform coordinated planning for a global action plan. The coordination of using emergency resource is a key issue of coordinated planning for an valid emergency response plan. In order to handle the real-time generated emergency resource conflict along with the distributed planning process, an hierarchical task network (HTN) planning process based multi-agent distributed coordinated planning method is designed. The method embeds the coordination process into hierarchical task network (HTN) planning process with the consideration that the action plan generation process should also be coordinated. This method can reduce the generation of useless plan and improve the efficiency of cooperation. Additionally, the coordination mechanism dealing with the reusable resources is developed, and a variety of priority rules are used to resolve resource using conflicts. Next, based on these designs, an hierarchical task network (HTN) based coordinated planning algorithm is proposed and implemented. Finally, an experimental study was presented to demonstrate the application of emergency resource coordinated planning and the validity and the efficiency of the method.

Influenza virus is one of major threats to human health. To explore effective strategies for prevention a novel influenza pandemic, the effectiveness of the prevention interventions including border quarantine, case treatment, case isolation, social-antiviral prophylaxis, and vaccination for four different influenza were computed based on the constructed computational epidemic model of Beijing. The main results were as follows: border quarantine could only play very limited role in preventing an influenza pandemic; for novel influenza virus with transmissibility similar to A(H1N1)/2009 virus, only positive treatment of cases could effectively prevent pandemic in local; for novel influenza virus with transmissibility similar to A(H2N2)/1958 virus, high intensity treatment of cases plus isolation of these treated cases could effectively prevent pandemic; for novel influenza virus with transmissibility similar to A(H1N1)/1918 virus, high intensity treatment of cases plus isolation of these treated cases and social-antiviral prophylaxis of treated cases could have a chance to prevent pandemic; for novel influenza virus with super transmissibility (R₀>3.3), only broad-spectrum vaccine with certain effectiveness could prevent pandemic. These results had significant value for further developing emergency preparedness plans to response a novel influenza pandemic in huge cities of China.

Foodborne diseases are usually underreported due to its various symptoms. It has emerged to be a critical burden of public health in China. Auto detection of foodborne disease event and risk assessment based on it are helpful to prevent and control its outbreak. We design three different event detection models according to three different kinds of data from disease surveillance system, food detection system and social media. By the integrated spatio-temporal data framework and the imported big-data of population, traffic, and food production and sales, the final model performs better than the isolated ones in spatial and temporal dimensions. This is testified by the results on the data of one city of China in 2014.