Yan Xinping

Evacuation Flow Assignment based on Improved MCMF Algorithm

During special events in urban transport network, traffic flow supported by route network is obviously different to that in normal circumstance. In the process of special events evacuation, itpsilas not only hoped a mass flow can be evacuated to some shelters, but also wished the evacuation time minimum. Therefore it is a problem to find a MinCost-MaxFlow in transport network. In this paper, the traditional MCMF was modified to describe network traffic assignment problems. Further more, a static traffic assignment based on pipelining was proposed to solve the evacuation flow problem. The numerical experiment on hypothetic network indicated the proposed MCMF algorithm can deal with evacuation traffic assignment problem reasonably.

Fault diagnosis for spark ignition engine based on multi-sensor data fusion

In data fusion approaches, Dempster-Shafer (D-S) evidence theory offers an interesting tool to combine data from multi-sensor. The decision-level fusion based on Dempster-Shafer (D-S) evidence theory can process non-commensurate data and has robust operational performance, reduces ambiguity, increases confidence, and improves system reliability. This paper describes mainly a decision-level data fusion technique for fault diagnosis for electronically controlled spark ignition engines. A D-S evidence theory fault diagnosis model is founded, and the feature selection and extraction of fault signal is conducted. Experiments on a 462 mini engine show that the data fusion technique provides good engine fault diagnosis method.

The Fusion Algorithm of Genetic and Ant Colony and Its Application

Genetic algorithm (GA) has strong adaptability, robustness and the quick global searching ability. It has such disadvantages as premature convergence, low convergence speed and so on. Ant colony algorithm (ACO) converges on the optimization path through pheromone accumulation and renewal. It has the ability of parallel processing and global searching and the characteristic of positive feedback. But the convergence speed of ACO is lower at the beginning for there is only little pheromone difference on the path at that time. The fusion algorithm of genetic and ant colony algorithm adopts genetic algorithm to give pheromone to distribute. And then it makes use of ant colony algorithm to give the precision of the solution. It develops enough advantage of the two algorithms. The comparative analysis on optimal performance of three algorithms is made by using the Camel function. Finally, the algorithm is used for the optimized the signal cycle length and green time by considering the constraint of automotive exhaust emission. The performance index function for optimization is defined to improve traffic quality and reduce emission at intersections. The simulation results show that very nice effects are obtained.

The Enhancement for Foggy Traffic Image Based on EM Algorithm

In bad weather, such as foggy, the vision become poor, and this often induce many problems for traffic safety. So it is a hot research subject how to improve the vision ability for bad weather in traffic, and there are many achievements have been made. Among these the single scale Retinex algorithm is a common used method, which was proposed by Land. According this theory, any obtained image can be divided into two images: illumination image and object reflected image. If the illumination can be estimated accurately, the high quality enhanced image (reflected image) will be obtained easily. In this paper, one efficient method (EFTI-EM) for illumination was proposed, which is used to estimate the illumination by using sky area in a foggy traffic image and to get sky area is based on EM algorithm. By setting a considerable number of tests, satisfactory results have been achieved.

QGA based Bandwidth Adaptation Scheme for Wireless/Mobile Networks

Radio resource management plays a major role in QoS provisioning for wireless/mobile networks. The scarcity and fluctuation of wireless link bandwidth, as well as the interconnection of wireless networks and Internet, motivate the research on adaptive multimedia services which have already run in Internet widely. The problem of seeking the optimal solution for revenue maximization is NP-hard, so the optimal bandwidth adaptation algorithm is not practical. In this paper, we propose a QGA (quantum genetic algorithm) based bandwidth adaptation scheme to find the suboptimal solution for such problem. As a general optimization algorithm, QGA is of less searching generations, higher convergence speed, higher searching efficiency, and better global searching ability, when compared with conventional genetic algorithm. Simulation results show that our algorithm works well for such problem, and with much less optimization time compared with the optimal algorithm

Tribological properties of DLC film prepared by C+ ion beam-assisted deposition (IBAD)

C+ ion beam-assisted deposition was utilized to prepare deposit diamond-like carbon (DLC) film. With the help of a series of experiments such as Raman spectroscopy, FT-IR spectroscopy, AFM and nanoindentation, the DLC film has been recognized as hydrogenated DLC film and its tribological properties have been evaluated. The ball-on-disc testing results show that the hardness and the tribological properties of the DLC film produced by C+ ion beam-assisted deposition are improved significantly. DLC film produced by C+ ion beam-assisted deposition is positive to have a prosperous tribological application in the near future.

Green port environment monitoring system with early warning and linkage control

Ports are rally points and hubs of water and land transportation, important nodes of the internet of things as well, which are inevitably susceptible to surrounding meteorological environment. To address this issue, a comprehensive environment monitoring system with early warning and linkage control is presented in this paper. First, our system can monitor surrounding meteorological environment data in real-time, which ensures to provide early warming if there encounters any exception. Second, our system can be used to forecast future environment, which significantly benefits the operations and schedules of ports. The last but not the least, through linkage control with other relevant systems, our system enables to meet the needs of green port development. The framework of our system, together with the design of some main software modules, are described in details. Furthermore, several main characteristics and functionalities of our system are further thoroughly introduced.

Remote fault diagnostic model for tribological systems in marine diesel engine with two-level self-organizing map network

Many researches indicate that a great number of failures occur in the tribological system which will reduce the reliability of the marine diesel engine. Therefore, it is necessary to monitor the condition and identify the fault mode of the engine. In this paper, remote fault diagnostic technology is developed to take full advantage of the online oil monitoring system and the laboratory analysis for the tribological systems. To increase the efficiency of fault diagnosis, a two-level fault diagnostic model based on self-organizing map (SOM) was established with the oil parameters from the online oil system and the experimental data from the laboratory. Based on the component map of SOM network, the attributes of the feature vector in the second level were reduced to simplify the model and the trajectory of the samples was tracked during the application of the system. The diagnostic result indicates that the remote fault diagnosis technology benefits the full acquirement of the information reflecting the engine condition and the two level fault diagnostic model can be well applied in fault diagnosis for the tribological systems in marine diesel engine with satisfactory result.

Development of ME-GI dual-fuel engine fault diagnosis expert system based on self-learning ontology

In order to meet intelligent requirement of the expert system in the process of fault diagnosis, a fault diagnosis system architecture which based on self-learning ontology was proposed in this paper. The fault diagnosis knowledge structure was defined; the relevant structure ontology and core fault ontology were constructed. Based on design of data warehouse of fault diagnosis, the decision tree in machine learning and Apriori algorithm were used to acquire fault knowledge to realize ontology self-learning. Taking the Hydraulic Control System of ME-GI dual-fuel engine as a prototype, the Hydraulic Control System diagnosis expert system was developed based on self-learning ontology.

Visibility estimated in foggy road traffic based on atmospheric scattering model

In road traffic, there exists foggy condition very often. The foggy affects traffic efficiency, even leads to traffic accidents. So it is important to measure the visibility in foggy road traffic.With this problem, two foggy estimated algorithms are proposed in this paper. In first algorithm, two iamges are used to estimate visibility togather with Atmospheric Scattering Model.The two image must be in same scene and different weather condition.The second algrithm is used one foggy image to estimate the visibility. But there are two key points in the scene,and these points have similar illuminate property.Experments show that these algrithms are effective.