Colonial bacterial memetic algorithm and its application on a darts playing robot
- Szilárd Kovács 1, Csaba Budai 2, János Botzheim 3
- Szilárd Kovács 1, Csaba Budai 2, János Botzheim 3
- 1Department of Artificial Intelligence, Faculty of Informatics, Eötvös Loránd University, Pázmány P. sétány 1/A, Budapest, Pest, 1117, Hungary. kovacsszilard@inf.elte.hu.
- 2Department of Mechatronics, Optics and Mechanical Engineering Informatics, Faculty of Mechanical Engineering, Budapest University of Technology and Economics, 4-6 Bertalan Lajos Street, Budapest, Pest, 1111, Hungary.
- 3Department of Artificial Intelligence, Faculty of Informatics, Eötvös Loránd University, Pázmány P. sétány 1/A, Budapest, Pest, 1117, Hungary.
- 0Department of Artificial Intelligence, Faculty of Informatics, Eötvös Loránd University, Pázmány P. sétány 1/A, Budapest, Pest, 1117, Hungary. kovacsszilard@inf.elte.hu.
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.The Colonial Bacterial Memetic Algorithm (CBMA) offers efficient robotic optimization. This advanced evolutionary approach excels in complex tasks, achieving high success rates and outperforming other methods.
Area Of Science
- Robotics
- Artificial Intelligence
- Evolutionary Computation
Background
- Robotic applications often face complex challenges including constraints, multiple objectives, and large search spaces.
- Existing optimization algorithms may struggle with efficiency and accuracy in these demanding scenarios.
- There is a need for advanced, adaptive optimization methods tailored for complex robotic tasks.
Purpose Of The Study
- To introduce the Colonial Bacterial Memetic Algorithm (CBMA) as an advanced evolutionary optimization approach for robotics.
- To enhance the Bacterial Memetic Algorithm by integrating Cultural Algorithms and bacterial group behavior-inspired co-evolutionary dynamics.
- To demonstrate CBMA's capability in handling complex robotic challenges, delivering fast and accurate solutions.
Main Methods
- CBMA integrates Cultural Algorithms and co-evolutionary dynamics with the Bacterial Memetic Algorithm.
- Features include multi-level clustering, dynamic gene selection, hierarchical population clustering, and adaptive co-evolutionary mechanisms.
- The algorithm was tested on a real-world robot arm ball-throwing task and the CEC-2017 benchmark suite.
Main Results
- Achieved a 100% success rate in a real-world robot arm ball-throwing task, often with fewer iterations and evaluations.
- Outperformed state-of-the-art algorithms on the CEC-2017 benchmark suite, showing superior outcomes in 71% of high-dimensional cases.
- Demonstrated up to an 80% reduction in required evaluations compared to other methods.
Conclusions
- CBMA is an efficient, adaptable, and robust evolutionary optimization algorithm for specialized robotic tasks.
- It effectively balances exploration and exploitation, offering significant advancements in adaptive evolutionary optimization for robotics.
- The algorithm's performance in both real-world and benchmark evaluations highlights its suitability for complex robotic applications.
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.