Presentation is loading. Please wait.

Presentation is loading. Please wait.

Speedy Agent Car: The Prototype Agent Technology: Final Project Lecturer: Prof. Ho Cheng-Seen Presented by: M9215801 - Irfan Subakti NTUST, June 7 th 2004.

Similar presentations


Presentation on theme: "Speedy Agent Car: The Prototype Agent Technology: Final Project Lecturer: Prof. Ho Cheng-Seen Presented by: M9215801 - Irfan Subakti NTUST, June 7 th 2004."— Presentation transcript:

1 Speedy Agent Car: The Prototype Agent Technology: Final Project Lecturer: Prof. Ho Cheng-Seen Presented by: M9215801 - Irfan Subakti NTUST, June 7 th 2004 Agent Technology: Final Project Lecturer: Prof. Ho Cheng-Seen Presented by: M9215801 - Irfan Subakti NTUST, June 7 th 2004

2 Overview Agent Implementation Agent Implementation Fuzzy sets theory on Agent Fuzzy sets theory on Agent Case study: GPS supported car  Speedy Agent Car’s prototype Case study: GPS supported car  Speedy Agent Car’s prototype Development possibilities Development possibilities Agent Implementation Agent Implementation Fuzzy sets theory on Agent Fuzzy sets theory on Agent Case study: GPS supported car  Speedy Agent Car’s prototype Case study: GPS supported car  Speedy Agent Car’s prototype Development possibilities Development possibilities

3 Purpose To show the implementation of agent To show the implementation of agent Here, this agent empowered fuzzy set theory Here, this agent empowered fuzzy set theory To show the implementation of agent To show the implementation of agent Here, this agent empowered fuzzy set theory Here, this agent empowered fuzzy set theory

4 Implementation of Agent Agent: A computing entity that performs user delegated tasks autonomously Agent: A computing entity that performs user delegated tasks autonomously Characteristics: Characteristics: Delegation Communication skills Autonomy Monitoring Actuation Intelligence We implemented agent in order to support user on maintain speed and distance between their cars. We implemented agent in order to support user on maintain speed and distance between their cars. Supported by fuzzy set theory and work autonomously, thinking by itself but can monitored as well. Also it delegated for most part activities, and communicate via GPS services. User can actuate this agent or just turn off. Supported by fuzzy set theory and work autonomously, thinking by itself but can monitored as well. Also it delegated for most part activities, and communicate via GPS services. User can actuate this agent or just turn off. Agent: A computing entity that performs user delegated tasks autonomously Agent: A computing entity that performs user delegated tasks autonomously Characteristics: Characteristics: Delegation Communication skills Autonomy Monitoring Actuation Intelligence We implemented agent in order to support user on maintain speed and distance between their cars. We implemented agent in order to support user on maintain speed and distance between their cars. Supported by fuzzy set theory and work autonomously, thinking by itself but can monitored as well. Also it delegated for most part activities, and communicate via GPS services. User can actuate this agent or just turn off. Supported by fuzzy set theory and work autonomously, thinking by itself but can monitored as well. Also it delegated for most part activities, and communicate via GPS services. User can actuate this agent or just turn off.

5 Knowledge of Speedy Agent Car Regarding fuzzy sets theory in Agent Technology subject, it’s interesting to implement some theory into real world (i.e., programming language). Regarding fuzzy sets theory in Agent Technology subject, it’s interesting to implement some theory into real world (i.e., programming language). Knowledge of Speedy Agent Car, defined by: Knowledge of Speedy Agent Car, defined by: IF Distance is Narrow AND Speed is Slow THEN Command=KeepSpeed IF Distance is Narrow AND Speed is Fast THEN Command=SlowDown IF Distance is Wide AND Speed is Slow THEN Command=SpeedUp IF Distance is Wide AND Speed is Fast THEN Command=KeepSpeed Regarding fuzzy sets theory in Agent Technology subject, it’s interesting to implement some theory into real world (i.e., programming language). Regarding fuzzy sets theory in Agent Technology subject, it’s interesting to implement some theory into real world (i.e., programming language). Knowledge of Speedy Agent Car, defined by: Knowledge of Speedy Agent Car, defined by: IF Distance is Narrow AND Speed is Slow THEN Command=KeepSpeed IF Distance is Narrow AND Speed is Fast THEN Command=SlowDown IF Distance is Wide AND Speed is Slow THEN Command=SpeedUp IF Distance is Wide AND Speed is Fast THEN Command=KeepSpeed

6 Knowledge of Speedy A. Car (cont’d) Explanation: Explanation: 'Narrow' are between 0 ~ 30 m 'Wide' are between 10 ~ more than 30 m 'Slow' are between 0 ~ 70 km/hours 'Fast' are between 30 ~ more than 70 km/hours All the max index set to value MaxIdx (e.g., 100) to make it convenient. Remember that index are between 0 ~ MaxIdx But the number of index is MaxIdx + 1 'KeepSpeed' are between -10 ~ 10 km/hours2 'SlowDown' are between -10 ~ 0 km/hours2 'SpeedUp' are between 0 ~ more than 10 m/hours2 Explanation: Explanation: 'Narrow' are between 0 ~ 30 m 'Wide' are between 10 ~ more than 30 m 'Slow' are between 0 ~ 70 km/hours 'Fast' are between 30 ~ more than 70 km/hours All the max index set to value MaxIdx (e.g., 100) to make it convenient. Remember that index are between 0 ~ MaxIdx But the number of index is MaxIdx + 1 'KeepSpeed' are between -10 ~ 10 km/hours2 'SlowDown' are between -10 ~ 0 km/hours2 'SpeedUp' are between 0 ~ more than 10 m/hours2

7 Algorithm Here, Car1 is in front of Car2. We are in Car2. Here, Car1 is in front of Car2. We are in Car2. All we have to do is maintain our car (Car2) regarding speed of Car1 and distance between our car and Car1. All we have to do is maintain our car (Car2) regarding speed of Car1 and distance between our car and Car1. We can trait our car to do something (i.e., command) whether keep our current speed, slow down with decelerate our speed or speed up with step on the gas. We can trait our car to do something (i.e., command) whether keep our current speed, slow down with decelerate our speed or speed up with step on the gas. Here, Car1 is in front of Car2. We are in Car2. Here, Car1 is in front of Car2. We are in Car2. All we have to do is maintain our car (Car2) regarding speed of Car1 and distance between our car and Car1. All we have to do is maintain our car (Car2) regarding speed of Car1 and distance between our car and Car1. We can trait our car to do something (i.e., command) whether keep our current speed, slow down with decelerate our speed or speed up with step on the gas. We can trait our car to do something (i.e., command) whether keep our current speed, slow down with decelerate our speed or speed up with step on the gas.

8 Algorithm (continued) Defined fuzzy sets of: Defined fuzzy sets of: Distance  Narrow & Wide Speed  Slow & Fast Command  KeepSpeed, SlowDown & SpeedUp Random acceleration of Car1, set speed of Car1 and update position of Car1. Random acceleration of Car1, set speed of Car1 and update position of Car1. Find distance between Car1 & Car2. Find distance between Car1 & Car2. From speed of Car1 and distance between two cars, apply 4 rules above to get command that Car2 have to do (whether KeepSpeed, SlowDown or SpeedUp). From speed of Car1 and distance between two cars, apply 4 rules above to get command that Car2 have to do (whether KeepSpeed, SlowDown or SpeedUp). Update graph and values during process, also position of Car2. Update graph and values during process, also position of Car2. Defined fuzzy sets of: Defined fuzzy sets of: Distance  Narrow & Wide Speed  Slow & Fast Command  KeepSpeed, SlowDown & SpeedUp Random acceleration of Car1, set speed of Car1 and update position of Car1. Random acceleration of Car1, set speed of Car1 and update position of Car1. Find distance between Car1 & Car2. Find distance between Car1 & Car2. From speed of Car1 and distance between two cars, apply 4 rules above to get command that Car2 have to do (whether KeepSpeed, SlowDown or SpeedUp). From speed of Car1 and distance between two cars, apply 4 rules above to get command that Car2 have to do (whether KeepSpeed, SlowDown or SpeedUp). Update graph and values during process, also position of Car2. Update graph and values during process, also position of Car2.

9 Graphs Below are graphs that depict fuzzy sets about Distance, Speed and Command. Below are graphs that depict fuzzy sets about Distance, Speed and Command.

10 Running Program Animation of Car1 & Car2. Car2 that always maintain it’s speed & distance to Car1. Animation of Car1 & Car2. Car2 that always maintain it’s speed & distance to Car1. Graph of Distance, Speed & Command. It come from knowledge that user defined. Graph of Distance, Speed & Command. It come from knowledge that user defined. Animation of graph obtained by program from rule #1, #2, #3, #4 & Final Result. Animation of graph obtained by program from rule #1, #2, #3, #4 & Final Result. Values obtained during running. It corresponds to Car1 & Car2 Values obtained during running. It corresponds to Car1 & Car2

11 Result and Evaluation From this prototype, we succeed to implement fuzzy set theory into an agent that maintenance speed and distance of a car, regarding with another car. From this prototype, we succeed to implement fuzzy set theory into an agent that maintenance speed and distance of a car, regarding with another car. This is an prototype, so we assumed that position and speed of another car obtained by GPS service. This is an prototype, so we assumed that position and speed of another car obtained by GPS service. Actually this is just simulation, for a real implementation, we have to consider: Actually this is just simulation, for a real implementation, we have to consider: GPS Hardware implementation Sophisticated coordination between involved parts Regulations from government Community supported Feasibility study by many parties From this prototype, we succeed to implement fuzzy set theory into an agent that maintenance speed and distance of a car, regarding with another car. From this prototype, we succeed to implement fuzzy set theory into an agent that maintenance speed and distance of a car, regarding with another car. This is an prototype, so we assumed that position and speed of another car obtained by GPS service. This is an prototype, so we assumed that position and speed of another car obtained by GPS service. Actually this is just simulation, for a real implementation, we have to consider: Actually this is just simulation, for a real implementation, we have to consider: GPS Hardware implementation Sophisticated coordination between involved parts Regulations from government Community supported Feasibility study by many parties

12 Development Possibilities Here we just concerning about the position of car in front or behind another car, we can develop to all possibilities of position on the road. Here we just concerning about the position of car in front or behind another car, we can develop to all possibilities of position on the road. Another things were possibly: Another things were possibly: Moving object on the road, not just cars. Obstacles on the road Traffic light arrangement Traffic jam resolution Distribution of flows on multiple path on the road Effective traveling path For common transportation: on the land, water or on the air Military system Here we just concerning about the position of car in front or behind another car, we can develop to all possibilities of position on the road. Here we just concerning about the position of car in front or behind another car, we can develop to all possibilities of position on the road. Another things were possibly: Another things were possibly: Moving object on the road, not just cars. Obstacles on the road Traffic light arrangement Traffic jam resolution Distribution of flows on multiple path on the road Effective traveling path For common transportation: on the land, water or on the air Military system

13 Thank You! For Your Attention …


Download ppt "Speedy Agent Car: The Prototype Agent Technology: Final Project Lecturer: Prof. Ho Cheng-Seen Presented by: M9215801 - Irfan Subakti NTUST, June 7 th 2004."

Similar presentations


Ads by Google