Research Of The Improvement Of TLBO Algorithms And It's Applications

With the development of society, swarm intelligence optimization algorithm has been widely developed and applied as a heuristic optimization algorithm. Due to the infinitized characters of optimization objective function such as continuation and differentiability, the algorithm shows the good applicability and quickly becomes a new research focus in optimization areas. In 2010, TLBO algorithm was presented by Rao. Compared to other swarm intelligence optimization algorithm, the TLBO algorithm needs few parameters to set, can be easily implemented and understood, solves quickly with high precision, and has good convergence capability as well.Firstly, the paper introduces what the TLBO algorithm is, simulates and compares the optimizing effect by applying TLBO algorithm and DE algorithm to seven common Benchmark functions. The result shows that TLBO algorithm is easy to fall into the local optimal solution in high-dimensional multi-peak test functions.Secondly, for the shortcomings in TLBO algorithm, TTLBO algorithm is designed to improve from three aspects: adaptive factor of teaching, increasing tacit understanding value M between teachers and students, and increasing weight of students 'self-confidence; moreover, the paper summarizes the previous improved TLBO algorithm, testing through several typical trial functions and making optimal comparison to TTLBO algorithm. The result shows that TTLBO algorithm has better indicators in optimal value, mean value, and standard deviation; meanwhile, applying TTLBO algorithm on three common constraint reference mechanical design and comparing with other three algorithms, the new algorithm shows a better optimization results for these three design problems.Finally, the paper describes what Adaptive Front Lighting System(AFS) is and put forward several adaptive control strategies to control front lighting systems. TTLBO algorithm can be applied to AFS system. Through calculating the overall satisfaction, we found that TTLBO algorithm's control strategy has better overall satisfaction than other control strategies as well as greatly improvement.