Objective1. Introduction2. The SystemFigure 4. The membership function of Traffic Flow/ Capability. The parameter will be maintained in the range 0~1.2.2 Fuzzy Inference System3. Conclusion & Future ResearchECE/CS 539 Final Project Shan-Huen HuangDecember 20, 2001OutLineAbstract IIObjective II 1. Introduction III 2.The System IV 2.1 Define Fuzzy Sets IV 2.1.1 Membership function of visibility IV 2.1.2 Membership function of Pavement Condition V 2.1.3 Membership function of Traffic flow/ Link capacity VI 2.1.4 Membership function of Speed Impact VII 2.2 Fuzzy Inference System VIII 2.3 Defuzzification IX 3. Conclusion & Future Research XReference XIFiguresFigure 1. The Fuzzy system. This figure is from the MATLAB fuzzy logic tool box and it indicates the relationship of inputs, outputs and the fuzzy inference.IVFigure 2. The membership function of Visibility. VFigure 3. The membership function of Pavement Condition. VIFigure 4. The membership function of Traffic Flow/ Capability. VIIFigure 5. The membership function of Speed Impact. VIIIFigure 6. The rule base. IXFigure 7. The fuzzy inference system and the output of speed impact after defuzzification.XApplication of fuzzy logic on traffic impacts from adverse weatherIECE/CS 539 Final Project Shan-Huen HuangDecember 20, 2001AbstractWe have developed an Expert System based on fuzzy logic to predict the impact of adverse weather on traffic, given the parameters of visibility, pavement condition, and traffic flow/link capacity. Fuzzy theory and fuzzy logic were used to establish the model. Fuzzy tool box build in MATLAB was used to implement the fuzzy theory. All the data includes visibility, pavement condition, and traffic flow could be obtained from Internet. There are tons of websites providing the weather forecast and the http://traffic.tann.net/ website provides the real time data. Minneapolis’s traffic data is used for analyzing. All the membership functions (fuzzy sets) could be generated by the historical data and expert experience. Based on the observed data and expert experience, we will define the fuzzy rules (fuzzy inference) and obtain the output. Here we will have a parameter, which is traffic flow/link capacity to maintain the link without congestion.ObjectiveThe impact of adverse weather to traffic flow has long been a problem for transportation management. Having the traffic impact of adverse weather helps us to evaluate the traffic condition and speed predicting. Problems have been encountered in cases where the effectiveness of traffic speed prediction in terms of heavy fog, flooding, heavy raining, snowing and strong winds. Precisely predicting the impact of adverse weathers would be helpful in traffic management, operations and so on. Moreover, the ITS (intelligent Transportation System) will also be improved by having those information.Fuzzy logic is suggested to predict the impact of adverse weathers. The advantage of using the fuzzy approach is that very little calculations need to be done and is also easier to use, once the fuzzy system has been established. MATLAB, along with the Fuzzy Logic toolbox is the software used to create the fuzzy inference system.In order to quantify how serious adverse weather affects traffic, we describe impact in term of traffic speed. Assuming traffic speed remains the same curve shape while there are without incidents or specific events or anything else that will largely affect traffic speed. In addition, the condition of congestion will not be considered here because other IIECE/CS 539 Final Project Shan-Huen HuangDecember 20, 2001factors will be involved with it. That means the only variable is weather condition. Based on the difference of weather condition, we can get the impact on traffic speed.1. IntroductionThere are two major parameters caused by adverse weather that will affect traffic speed: visibility and pavement condition. Those data are sometimes not crisp, that means there are not well defined and clear. For instance, probably we will hear that “the visibility is good or bad” or “the pavement is slippery or a little bit slippery”. Those linguistic descriptions are vague and not easy to interpret into crisp definition. Therefore, fuzzy logic is adopted because of its easily modeling and calculating.This project is divided into 3 sections. Section one gives some overview about Fuzzy logic especially the mamdani model, Section 2 explains the system that was developed, Section 3 describe the conclusion and future research.Fuzzy Logic- Dr. Lotfi Zadeh of University of California, Berkeley, introduced fuzzy logic in the 1960s’. Fuzzy logic is a superset of conventional (Boolean) logic that has been extended to handle the concept of partial truth – truth-values between “completely true” and “completely false”. In fuzzy logic the truth of any statement becomes a matter of degree.The Mamdani Model- The Mamdani fuzzy model, similar to the other (Sugeno and Tsukomoto model) fuzzy models, has a set of inputs, each with its own membership functions. The Mamdani model differs from the others in the fact that the output also has a set of its own membership functions. In the case of the Sugeno model, the output is a function of the input membership function and in the case of Tsukomoto’s model; the output is a linear function of the input.Mamdani’s Method: VUBAcvuvuBAR ),/()()(IIIECE/CS 539 Final Project Shan-Huen HuangDecember 20, 2001)()(),( vuvuBABA where ""can be Min. or Product.2. The SystemIn order to set up the whole fuzzy system, there are three things: 1. Define fuzzy sets (Membership Functions).2. Set up the fuzzy inference system (rule base).3. Defuzzifier. After those procedures, we defuzified the traffic impact fuzzy sets and obtain a crisp speed impact as output. The output indicates the speed reduction under the adverse weather in a link with specific traffic flow. Figure1 shows the relationships of the system.Figure 1. The Fuzzy system. This figure is from the MATLAB fuzzy logic tool boxand it indicates the relationship of inputs, outputs and the fuzzy inference.2.1
View Full Document
Unlocking...