Fuzzy Logic & Knowledge-Based Systems (AI)
Fuzzy Logic & Knowledge-Based Systems (AI)
Fuzzy Logic & Knowledge-Based Systems (AI)
Module:
Module:
Module:
Fuzzy Logic & Knowledge-Based Systems (AI) IMAT3406
Fuzzy Logic & Knowledge-Based Systems (AI) IMAT3406
Fuzzy Logic & Knowledge-Based Systems (AI) IMAT3406
Project Duration:
Project Duration:
Project Duration:
2 Months
2 Months
2 Months
Year:
Year:
Year:
2024
2024
2024
Tool:
Tool:
Tool:
MatLab
MatLab
MatLab
Overview
This project focuses on designing a Fuzzy Inference System (FIS) for an autonomous maintenance machine operating in space. The goal was to create a system capable of handling the complexities of space missions, such as imprecise measurements, uncertain information, and nonlinear systems. The FIS evaluates inputs like distance to the satellite, relative velocity, altitude, fuel levels, and battery status to make decisions about maintenance actions such as rendezvous, repair, capture, or abort. The system was developed to enhance spacecraft performance, improve reliability, and increase autonomy in orbital operations. The project involved designing two versions of the FIS, with the second version (System V2) simplifying the inputs and outputs to improve interpretability and efficiency.
This project focuses on designing a Fuzzy Inference System (FIS) for an autonomous maintenance machine operating in space. The goal was to create a system capable of handling the complexities of space missions, such as imprecise measurements, uncertain information, and nonlinear systems. The FIS evaluates inputs like distance to the satellite, relative velocity, altitude, fuel levels, and battery status to make decisions about maintenance actions such as rendezvous, repair, capture, or abort. The system was developed to enhance spacecraft performance, improve reliability, and increase autonomy in orbital operations. The project involved designing two versions of the FIS, with the second version (System V2) simplifying the inputs and outputs to improve interpretability and efficiency.
This project focuses on designing a Fuzzy Inference System (FIS) for an autonomous maintenance machine operating in space. The goal was to create a system capable of handling the complexities of space missions, such as imprecise measurements, uncertain information, and nonlinear systems. The FIS evaluates inputs like distance to the satellite, relative velocity, altitude, fuel levels, and battery status to make decisions about maintenance actions such as rendezvous, repair, capture, or abort. The system was developed to enhance spacecraft performance, improve reliability, and increase autonomy in orbital operations. The project involved designing two versions of the FIS, with the second version (System V2) simplifying the inputs and outputs to improve interpretability and efficiency.
Challenges
The project faced several challenges, primarily due to the complexity of space operations. The initial version of the FIS (System V1) had 14 inputs and 7 outputs, making the decision-making process overly complicated and difficult to interpret. During testing, several critical rules failed to activate as intended, raising concerns about the system's reliability. Additionally, the system struggled with redundant variables and unclear outputs, which hindered its ability to make efficient decisions. The dynamic and unpredictable nature of space missions also posed challenges, as the system needed to handle imprecise data and uncertain conditions effectively.
The project faced several challenges, primarily due to the complexity of space operations. The initial version of the FIS (System V1) had 14 inputs and 7 outputs, making the decision-making process overly complicated and difficult to interpret. During testing, several critical rules failed to activate as intended, raising concerns about the system's reliability. Additionally, the system struggled with redundant variables and unclear outputs, which hindered its ability to make efficient decisions. The dynamic and unpredictable nature of space missions also posed challenges, as the system needed to handle imprecise data and uncertain conditions effectively.
The project faced several challenges, primarily due to the complexity of space operations. The initial version of the FIS (System V1) had 14 inputs and 7 outputs, making the decision-making process overly complicated and difficult to interpret. During testing, several critical rules failed to activate as intended, raising concerns about the system's reliability. Additionally, the system struggled with redundant variables and unclear outputs, which hindered its ability to make efficient decisions. The dynamic and unpredictable nature of space missions also posed challenges, as the system needed to handle imprecise data and uncertain conditions effectively.
Approach
To address these challenges, the project was divided into two phases. In the first phase, System V1 was designed with a comprehensive set of inputs and outputs, including distance to the satellite, relative velocity, altitude, battery level, fuel level, and target state. The system used fuzzy logic to handle imprecise data and nonlinear systems, allowing it to make informed decisions under uncertainty. However, after testing revealed issues with rule activation and interpretability, a second phase was initiated. In System V2, the number of inputs was reduced from 14 to 6, and the outputs were consolidated into three key decisions: thrust, reverse thrust, and a single "Decision" variable that combined actions like rendezvous, repair, capture, and abort. Advanced membership functions, such as Gaussian and Generalized Bell-shaped functions, were introduced to improve the system's ability to interpret linguistic variables and make nuanced decisions.
To address these challenges, the project was divided into two phases. In the first phase, System V1 was designed with a comprehensive set of inputs and outputs, including distance to the satellite, relative velocity, altitude, battery level, fuel level, and target state. The system used fuzzy logic to handle imprecise data and nonlinear systems, allowing it to make informed decisions under uncertainty. However, after testing revealed issues with rule activation and interpretability, a second phase was initiated. In System V2, the number of inputs was reduced from 14 to 6, and the outputs were consolidated into three key decisions: thrust, reverse thrust, and a single "Decision" variable that combined actions like rendezvous, repair, capture, and abort. Advanced membership functions, such as Gaussian and Generalized Bell-shaped functions, were introduced to improve the system's ability to interpret linguistic variables and make nuanced decisions.
To address these challenges, the project was divided into two phases. In the first phase, System V1 was designed with a comprehensive set of inputs and outputs, including distance to the satellite, relative velocity, altitude, battery level, fuel level, and target state. The system used fuzzy logic to handle imprecise data and nonlinear systems, allowing it to make informed decisions under uncertainty. However, after testing revealed issues with rule activation and interpretability, a second phase was initiated. In System V2, the number of inputs was reduced from 14 to 6, and the outputs were consolidated into three key decisions: thrust, reverse thrust, and a single "Decision" variable that combined actions like rendezvous, repair, capture, and abort. Advanced membership functions, such as Gaussian and Generalized Bell-shaped functions, were introduced to improve the system's ability to interpret linguistic variables and make nuanced decisions.
Solutions
To overcome the challenges, several key changes were implemented. In System V2, redundant inputs like distance to the satellite, target detection sensor, engine status, robotic arm status, and target lock were removed to streamline the decision-making process. The outputs were consolidated into a single "Decision" variable, making the system's responses more interpretable and easier to manage. Advanced membership functions were introduced to better represent linguistic variables, allowing the system to handle imprecise data more effectively. These changes significantly improved the system's reliability, interpretability, and adaptability, making it more suitable for the dynamic and uncertain environment of space missions.
Outcome
The project successfully delivered a functional FIS for autonomous maintenance in space. System V2, with its simplified inputs and outputs, proved to be more reliable and efficient than the initial version. The system's ability to handle imprecise data and make informed decisions under uncertainty was significantly enhanced, making it a valuable tool for space missions. The introduction of advanced membership functions improved the system's linguistic representation, allowing it to capture nuanced information and make contextually aware decisions. While the initial version faced challenges with rule activation and complexity, the refined System V2 addressed these issues and laid the foundation for future improvements. Overall, the project demonstrated the potential of fuzzy logic to enhance spacecraft performance, reliability, and autonomy in the challenging environment of space.
The project successfully delivered a functional FIS for autonomous maintenance in space. System V2, with its simplified inputs and outputs, proved to be more reliable and efficient than the initial version. The system's ability to handle imprecise data and make informed decisions under uncertainty was significantly enhanced, making it a valuable tool for space missions. The introduction of advanced membership functions improved the system's linguistic representation, allowing it to capture nuanced information and make contextually aware decisions. While the initial version faced challenges with rule activation and complexity, the refined System V2 addressed these issues and laid the foundation for future improvements. Overall, the project demonstrated the potential of fuzzy logic to enhance spacecraft performance, reliability, and autonomy in the challenging environment of space.
The project successfully delivered a functional FIS for autonomous maintenance in space. System V2, with its simplified inputs and outputs, proved to be more reliable and efficient than the initial version. The system's ability to handle imprecise data and make informed decisions under uncertainty was significantly enhanced, making it a valuable tool for space missions. The introduction of advanced membership functions improved the system's linguistic representation, allowing it to capture nuanced information and make contextually aware decisions. While the initial version faced challenges with rule activation and complexity, the refined System V2 addressed these issues and laid the foundation for future improvements. Overall, the project demonstrated the potential of fuzzy logic to enhance spacecraft performance, reliability, and autonomy in the challenging environment of space.