Parameters Identification Of Ship Longitudinal Motion Based On Improved Firefly Algorithm

During the process of ships traveling at sea,periodic oscillations may occur when they are affected by adverse sea conditions,which can bring adverse effects to the daily life,work,and even ship cargo of the crew.In order to avoid the impact of the marine environment on ship transportation,it is necessary to establish an accurate mathematical model of ship motion.Hydrodynamic parameters are an important part of building ship mathematical models,but some of the more sensitive hydrodynamic parameters are difficult to obtain,so this thesis uses the improved Firefly algorithm based on Roulette Wheel Selection Strategy(CRSFA)to identify the parameters of ship longitudinal motion.The Firefly algorithm(FA)is a relatively new swarm intelligence algorithm obtained by simplifying the behavior of communication between groups of fireflies relying on luminescence.This algorithm has the advantages of simple structure,fewer parameters that need to be adjusted,fast optimization accuracy and convergence speed.This thesis studies the basic theory of Firefly algorithm,introduces sigmoid function,roulette wheel strategy and chaos theory,improves the Firefly algorithm,and analyzes its application in the identification of ship longitudinal motion parameters.First,model the longitudinal motion of the ship and establish the differential equation of the longitudinal motion of the ship.Considering that the computer cannot process continuous signals and the collected data is also discrete signals,the differential equation of the longitudinal motion of the ship is discretization to facilitate the next step of work.And the Firefly algorithm to learn,analyze its defects,Firefly algorithm for the defects of improved methods.Secondly,the Firefly algorithm is improved.Considering the impact of the location update formula on the accuracy of the algorithm,the sigmoid function is introduced to improve the algorithm,and the sigmoid function is correspondingly changed for the Firefly algorithm,so that the location update formula of the Firefly algorithm becomes the location update formula of adaptive weight,avoiding the problem of repeated oscillations in the later iteration,and accelerating the convergence accuracy of the algorithm.The roulette wheel strategy and chaos theory are introduced to generate individuals with randomness and better fitness value to replace the inferior individuals in the population in the iterative process,forming a firefly population with better fitness,avoiding the algorithm from falling into local optimization,and improving the global search ability of the algorithm.For the improved algorithm,representative unimodal and multimodal test functions were used to optimize the algorithm through different dimensions,and the improved algorithm was compared with the original algorithm.The results show that the solution accuracy and convergence speed of CRSFA are better than the original algorithm,and it can avoid the problem of repeated oscillations in the algorithm,and can be effectively applied to ship longitudinal parameter identification.And identify the responsive model,and verify the feasibility of applying the improved algorithm to ship longitudinal parameter identification through the comparison of ship maneuverability index.Finally,the CRSFA parameter identification scheme for the longitudinal motion of the ship is designed,the appropriate fitness function is selected,the input signal is determined,the mathematical model of the longitudinal motion of the ship is established,the parameters of the HD702 boat with known hydrodynamic parameters are identified,and the HD702 boat with different heading is identified using the improved algorithm and the initial algorithm,and the results are compared and analyzed.Parameter identification was carried out on the KCS ship model,respectively under calm water and level five sea conditions.The identified parameters were used to predict the longitudinal motion attitude of the ship,and the motion trend almost coincided with the actual value.The feasibility of the identification scheme proposed in this thesis was verified by identifying the longitudinal motion parameters of different ships.