The Particle Swarm Optimization Algorithm Research And Application In Geotextiles-laying Vessel Intelligent Control

Geotextiles-laying vessel is a special-purpose project vessel which meet the need of Changjiang River channel renovate project,guarantee the renovate project go on smoothly.The "YUGONGPAI NO1"geotextiles-laying vessel has been the first vessel with the higher automatization monitoring system in region of the middle and lower reaches of the Yangtze River.The "CHANG YAN NO 1" and "CHANG YAN NO 2" desiged recently are both greater than it in the scale and operation.Automatic operation has become the development trend of Geotextiles-laying vessels' design, but due to the needs of the development of waterway dredging industry,the desiged vessels became large,with more number of winches,working mode also become more complicated,the design of automatic control for it becomes more complex,the original design of simple fuzzy control strategy has not meet the design requirements.it is necessary to optimize the original control strategy of the vessel.In this paper,based on the deeply analysis and study of the convergence theory and several improvement methods of PSO algorithm,a new QPSO with two Particle Swarms based on public history researching side-by-side(TPHQPSO) is presented,in order to strengthen the convergence speed and global searching ability of PSO. Comparative experiments on testing functions indicate that this algorithm improves convergence speed and enhances the global searching ability.Through analyzing the characteristics of the current automatic control system for Geotextiles-laying vessel and working condition,this paper constructs a network structure of automatic control system for Geotextiles-laying vessel with more winches. Aiming at accurately controlling the vessels' movement,an adaptive fuzzy control scheme based on neural network modeling is proposed.In this scheme,control model is constructed using BP neural network with real-time working data as the training sample.the neural network model for Geotextiles-laying vessel is optimized with online learning.According to the deviation between the output of neural network model and the ideal input trajectory,the parameters of the fuzzy controller is real-time adjusted still using the TPHQPSO algorithm proposed in this paper.Based on the MATLAB software environment,the control scheme for removing vessel is simulated experimental.and the results show that the designed controller can eliminate most deviation,and able to track the changes of deviation,the adaptive fuzzy control scheme meets design requirements.Finally,current design automatization monitoring system for "CHANG YAN NO 2" geotextiles-laying vessel is simplely introduced in this paper,including function and interface of the monitoring system,design method and application of automatic control program for the Geotextiles-laying vessel.Experiments show that,the fuzzy adaptive control scheme for remove the vessel, based on proposed "TPHQPSO" optimization algorithm,can solve the problem of six anchor linkag the boat moved ship chessboard in a certain extent can solve the problem of six anchor linkage in a certain extent.But due to the complicated working condition of the vessel now,only in the downstream condition this method works well,and the method under the other conditions is still under further research.