Haibin Yu

Auction-based Dynamic Coalition for Single Target Tracking in Wireless Sensor Networks

Wireless sensor networks have recently gained a lot of attention. Target tracking is a canonical application of wireless sensor networks and task allocation is an unavoidable problem for target tracking. To improve the tracking quality and save more energy, an auction-based dynamic coalition scheme of task allocation for single target tracking in wireless sensor networks was introduced. Furthermore, the scheme was extended to multiple dynamic coalition scheme to achieve even better performance. The simulations compared the standard deviation of tracking error, detecting ratio and energy consumption of the new schemes with those of case-based reasoning dynamic coalition scheme. The auction-based dynamic coalition scheme saves more energy and improves the tracking quality evidently. The extended scheme achieves better tracking quality and the energy consumed is still far less than that consumed in the case-based reasoning dynamic coalition scheme

Three-phase Induction Motor Operation Trend Prediction Using Support Vector Regression for Condition-based Maintenance

Due to the broad employment and large amount of electricity consumption of induction motor, their efficient operation has been a focus for engineering research. The paper proposes a new integrated approach performing the motor condition prediction for the maintenance of low cost and high quality. Studies were done on nonlinear data analysis techniques, including particle filters for state estimates and support vector regression for condition prediction. Laboratory studies support the condition-based maintenance for motor systems

Implementation of hybrid wired and Wireless Foundation Fieldbus

Wireless foundation fieldbus (WFF) is put forward and used to provide wireless communication services for process industrial field devices to implement monitoring, control and management tasks. WFF communication model consists of the foundation fieldbus (FF) user layer and application layer, the data link layer and the IEEE802.11 physical layer. The WFF data link layer combines FF and IEEE802.11 mechanisms to satisfy the requirements of the automation tasks. WFF can be hybrid with wired FF. The hybrid is implemented on the interoperability of WFF and FF user applications and devices. To guarantee the performance of a hybrid field network, a unified system management and scheduling mechanisms are used to construct and perform the automation tasks, and interconnecting devices convert and forward communications between FF and WFF devices on the data link layer. The analysis shows the communication performance of the hybrid network can satisfy the different real-time requirements of the automation system.

A distributed and optimal algorithm to coordinate the motion of multiple mobile robots

This paper addresses the problem of planning paths for multiple mobile robots moving in the same environments. We try to decrease the computational complexity of path planning, which is one of the fundamental limiting characteristics in the existing work of path planning. Unlike many other approaches to this problem, we assume the robots are restricted by the limited-communication range such that information about other robots can be known only when they move close to each other. In our method, the computationally expensive problem is decomposed into two modules-path planning and velocity planning. Such decomposition makes approaches of path planning for multiple mobile robots applicable. Trajectory for each robot is achieved by using the approach named vector field histogram (VFH) before the robots begin to move. Then the conflicts among robots are solved by adjusting the velocities of robots. Velocity planning occurs only inside the robots that are in the same network. The method of planning velocities for the robots is called the dynamic priority assignment, which can minimize the total waiting time of robots in the same network. Finally, both simulated and real-robot experiments have validated our approach.

Research of control system platform technology for IC equipment

With the development of road to 300 mm semiconductor wafer manufacturing, IC equipment control system not only implement its task scheduling, material transfer and process control, we must also consider the enterprise information integration, production scheduling and industry standards. This paper analyzes the semiconductor manufacturing process, control requirements, and presents a framework-style, modular IC equipment control system platform model, aimed at the complete system control functions at the same time, making the control of the semiconductor equipment greater flexibility, scalability, enterprise-class interactive capabilities, and in accordance with industry standards.

An optimized aggregators selection problem for industrial Wireless Sensor Networks

In-network data aggregation can reduce the data collection cost and improve the performance of wireless sensor networks (WSNs). Whereas there is no consideration for this in-network processing in current WSN protocols based on industrial application. Clustering sensor nodes is an effective technique to achieve in-network aggregation in WSNs. The goal of this paper is to demonstrate how judicious selection of a few aggregators from the sensor nodes can significantly reduce the data collection cost. The problem resembles the classical multi-objective network location problem. For large network, this paper shows an approximate algorithm which provides a suboptimal solution. It makes the selection of aggregators well-adapted to the network structure under the constraint of minimal data collection cost. Simulation results show that the algorithm efficiently reduces the data-collection cost compared with the un-aggregation data collection and enhances the transmission reliability.

Dynamic production scheduling system based on data warehouse

In todays manufacturing systems, the production-scheduling system in shops usually is a weak partition. The paper proposes an architecture of dynamic scheduling and presents the implementation structure of the dynamic production scheduling system based on data warehouse and knowledge discover in database. Then, the key technologies used in this system are discussed. The paper presents a new method of developing a scheduling system.