Jia Hongfei

Research on agent-based approaches to freight transport scheduling

In the last few years, the agent and agent-based technology have been researched and applied in a number of domains. However, the advantages and effectiveness are still under testing on freight transport scheduling. The aim of this paper is to examine the potential applicability of the agent-based technology to freight transport. The research started with literatures review, a real world experiment was then conducted by using an agent-based scheduling tool developed by Magenta. Two scenarios were tested in the experiments and the results were analyzed and evaluated by comparing with the actual operational performance. The results of the simple scenario were as good as the actual performance, while the complex scenario improved a lot. It is concluded from the study that there is an increasing trend and a number of positive aspects of the application of agent-based technology to the freight transport scheduling, the agent-based technology is suitable for solving the distributed, complex and volatile scheduling problems in the freight transport domain.

Adaptive-filtering based dynamic OD matrix estimation

This paper proposes one method for the dynamic origin-destination (OD) estimate. Based on the improved Sage-Husa adaptive filtering algorithm which carries on recursive filters observation data obtained by time-variable noise statistics estimator, estimate and revision observation noise statistical property, one can suppresses the divergence of the filter and, increase the filter precision. The experiment results indicated that this kind of method has the good performance.

Travel-time estimation based on federated filtering

Propose one method for the travel time estimate based the federated filtering. The federated filter was used to fuse the average travel measured by the fixed detectors and vehicle detectors to estimate the average travel time of the road. The fusion method is in favor of practical application because of its simple calculations and less time-consuming. The result shows that this method can improve the precision of average travel time estimation.

Dynamic order strategy(R, s t , S T ) for automobile spare parts on a target service level

This paper studies on dynamic order strategy for auto spare parts on a target service level. A new three-dimensional classification is proposed integrated order policy of spare parts. It improves the present the weakness of ABC classification and combines with three control variables influenced the order policy, X demand, Y item price and Z variation of demand, to classify all spare parts. Unlike most extant inventory-control approaches, Strategy(R, st, ST) offered is a kind of dynamic order strategy integrated with inventory control parameters with inventory check cycle (R), dynamic reorder point (st) and dynamic order quantity (ST) and is set the goal with a target service level while minimizing of total inventory cost. It takes into accounts the dynamic requirements in category, time and quantity in accordance with automobile spare parts. Simulation experiments were performed by ARENA 10.0 to represent the performance of strategy (R, st, ST) and the optimized strategy is given. Outcomes simulating show that dynamic strategy (R, st, ST) enables to carry out dynamic inventory management and control by parameters R, st, ST for automobile spare parts. Also a goal under a certain target service level while satisfied the minimization of total inventory cost can be come true.

IDENTIFYING AND ELIMINATING CONSTRAINTS IN REVERSE LOGISTICS SYSTEM OF END-OF-LIFE HOME APPLIANCE

This research is conducted mainly to focus on constraints in the reverse logistics system of End-of-Life home appliance in China. With case study of Remanufacturing Plant, Resource Constraints in system are identified by logistics simulation, and then they are eliminated finally through using Drum-Buffer-Rope (DBR) scheduling of Theory of Constraints. Inventory Buffer is set just before the constraint unit as well as buffer size model is established. Outcomes suggests that DBR scheduling is exactly an effective approach to eliminate Resource Constraints and it enables to achieve the objective of optimizing parameters of throughput (T), investment (I) and operating expense (OE), even logistics balance.

MEASURING THE EXTERNAL COSTS OF URBAN TRAFFIC CONGESTION

Qualitative and quantitative analyses of the external costs which urban traffic congestion brings are the base of improving travel demand management and designing relative strategies, and even will help to maximize the effective usage of urban road resources. Firstly, this paper analyses the production mechanism of urban traffic congestion: travelers overlook the negative externality of urban traffic congestion then join in the congested queues. The negative externality is obvious, since the explosion of vehicles in finite spaces and the augmentation of traffic densities arouse increases in travel time, traffic accidents, environmental pollution, and fuel consumption. Consequently, the external costs of traffic congestion are divided into four parts. We make an intensive study of the quantitative measurements for those four external costs. The extra travel time costs can be calculated by regional average time value. The environmental pollution costs can be measured by the proportion of congestion pollution in total transport pollution. The traffic accident costs can be figured out by the probability of accidents. The fuel consumption costs can be work out by the proportion of travel time. Also, a sample case is introduced to prove the feasibility of this method. It supplies a basis for the implement of urban road congestion pricing and other travel demand management strategies.