Research On Ship Course-control Based On Modified Particle Swarm Optimization

With the development of the Marine economy, the requirement of the safety andeconomy on the voyage is higher than ever before. At the same time,because of theincreasing of ship quantity, tonnage, shipping density, the voyage the complexity of theenvironment and uncertainty, the ship motion control also becomes more complicated.The ways of traditional ship course control algorithm include PID control and adaptivecontrol, etc. The ship course control mostly uses PID controller, and a small amount of shipsuse adaptive control etc. PID controller parameter setting depends too much on the priorknowledge and object model, consequently leading to overshoot, the regulating time long,steady-state error and other problems. Since cost of the adaptive control method is high andits parameters adjustment is also difficult, and the system is not stable. Also the heading ofcontrol method present have their applied limitations. Due to the ship maneuvering motionmathematical model of nonlinear, operation actuators lag and navigation environment and byinterference problems such as uncertainty, make some of the control method based on modelhardly to work. With the development of computer and modern control theory, new kinds ofintelligent control algorithms arise. These intelligent algorithms have been used on shipheading control system, and made some effect, but there are some problems because of modelinappropriate or the algorithm itself. So need of building a more appropriate simulationmodelu is urgent, and so is the study of more intelligent control theory applied to the controlof the ship course, so as to develop a high-performance heading control products.According to the analysis and simulation of the bird flock foraging process, Kennedy andEberhart in1995put forward the original Particle Swarm algorithm (Particle SwarmOptimization, PSO). And particle swarm algorithm quickly became evolutionary computationfield and intelligent optimization research’s hot spot, which has been widely used in datamining, and dynamic environment optimization, the objective function optimization, thefuzzy control system and the network training god, etc. It can be efficient, parallel searchproblem space of the corresponding coding space, and get a global optimal value. And it isfast and its math expression is simple. But it also have easy to early maturity and easy to alocal optimal value prombles. The study found that the particle swarm sport is regular, but theinitialization of whole population is random and unpredictable. If the particle swarmoptimization algorithm and the ergodicity of chaotic variables are combined, and then theergodicity of the chaos and initialization of the sensitivity will be given full scope. Theparticle will be more evenly spread in the whole search space, making the calculate method better in keeping population diversity, escaping local optimal value, improving the globalsearch ability algorithm and the success rate of convergence.The mathematical model of the ship course control system is set up in this paper. Thebasic Particle Swarm optimization algorithm and modified particle swarm optimizationalgorithm mentioned above are applied into optimization of PID controller of the ship coursecontrol system. After simulation and comparison, the results show that the PID controllerdesigned based on modified particle swarm optimization algorithm has better dynamic andstationary performance, and both the theoretic analysis and the simulation prove its validity.