Research Of Cloud Model Particle Swarm Optimization Algorithm And Its Application In Ship Engineering

In practical engineering applications and scientific research, many of the problemshave difficulties with binding, complexity, multiple local minima, nonlinear modeling.Finding a suitable optimization algorithm is an important research direction. ParticleSwarm Optimization (PSO) is an evolutionary algorithm proposed by Kennedy andEberhart in1995, which originated from the simulation of the behavior of bird predation.The algorithm can be used to solve complicated optimization problem without requiringcentralized control and global modeling.The main research works are discussed as follows.(1) Cloud model theory is described in detail, including the basic concepts ofmathematical characteristics and the statistical analysis of the cloud model.(2) A cloud adaptive particle swarm optimization algorithm based on cloud variation isproposed. For the ratio of the global optimal value and particle fitness reflects thecharacteristics of the particle optimal difference, use cloud model generator to adaptivelyadjust the inertia value of every particle. Use normal clouds operator to realize variationoperation of part particle, and set the parameters of algorithm correctly. Simulation resultsof typical test functions shows that the proposed algorithm has fine capability of findingglobal optimum, fast convergence, and high accuracy. It is suitable for find the value ofmulti-modal function. This algorithm is a practical and effective method.(3) The vertical motion of the ship hydrodynamic model is studied. The improvedadaptive cloud particle swarm optimization algorithm is used to design a vertical motion ofthe ship hydrodynamic parameter identification. The simulation results show that thisalgorithm has good stability, high accuracy of recognition. And it can meet the actualdemand in solving ship vertical motion of hydrodynamic parameter identification problem.(4) The PID control for ship steering problem is studied. The adaptive cloud particleswarm optimization algorithm is applied to optimize the three parameters of PID controller.The simulation results show that system performance has been greatly improved, includingwithout overshoot, rise fast, stable, strong robustness.