Kong Yaguang

Urban Traffic Incident Detection Based On Fuzzy Logic

Automatic incident detection and characterization is urgently required in the development of advanced technologies used for reducing non-recurrent traffic congestion on urban traffic. This paper presents a new method which is constructed on the basis of fuzzy value logic to automatically identify freeway incidents. As a component of a real-time traffic adaptive control system for signal control, the algorithm feeds an incident report to the systems optimization manager, which uses the information to determine the appropriate signal control strategy. Off-line tests were conducted to substantiate the performance of the proposed incident detection algorithm based on simulated data. The test results indicate the feasibility of achieving real-time incident detection utilizing the proposed method

Design of Real Time Control Software Based On QNX

System architecture of one real time control software is given in this article, and the design method is introduced for each module. The real time control software support the redundant of system network, control network and control station, enhance the robustness of system greatly; select QNX as real time operation system, support control algorithm editor based on IEC61131-3, and tasks with different period or stimulate by event all supported in this software; receive operate instruction and unload real time value through Ethernet by socket; in the handle program of interrupt from CAN, acquisition of real time data from 32 IO board can be finished in 320 ms. This real time control software has been used in one DCS system, and has strong robustness and high real time ability

Model Identification of the Continuous Casting Billet Induction Heating Process for Hot Rolling

This paper firstly analyzes the continuous casting billet induction heating process in hot rolling line to build an ARX model. Then, it determines the model order by AIC criterion, and estimates parameters by the least square method to realize ARX model identification. Finally, the identified model is verified by an actual example, and the result shows that the model has high accuracy and can provide the basis for next temperature control.