Concise Robust Optimization Control Of Track-keeping Autopilot With Rudder Roll Stabilization

Ship roll stabilization is a very important research trend in ship motion control field. Rudder roll stabilization (RRS) is a new ship roll stabilization device based on the traditional course autopilot system. The autopilot with RRS can achieve good roll reduction rate and satisfactory course-keeping performance, and overcome the disadvantages of stabilizing fins, and then improve the ship’s seaworthiness, navigational safety and personnel comfort in rough sea conditions.The thesis has made an intensive study of track keeping autopilot with RRS. Firstly, analysis is done on the underactuation of ship motion and the non-minimum phase characteristic of roll responding, and the coupling mechanism between yaw and roll in case of big amplitude rudder is determinated qualitatively and quantitatively by simulations. The results indicate that the coupling mechanism between yaw and roll is obvious, and the amplitudes of yaw and roll have strong positive correlation property.Based on above analysis, to establish a more precise ship responding motion model, a roll responding model is brought up based on the full consideration of non-minimum phase, damped oscillation and inertial heeling of the roll responding to rudder motion. The model parameters have obvious physical meanings, which can be estimated simply according to the responding curve. The rudder responding outputs of a nonlinear multifunctional naval vessel model are set as the test data. The parameters of the ship responding motion model are identified by adaptive GA algorithm, and then the yaw and roll responding models are constructed. Simulation tests results indicate that the ship responding motion model can approach the test data with little error, which verifies the validity and reliability of the ship responding motion modeling.The closed-loop gain shaping algorithm is a simplified H∞robust control theory, which constructs the controller using four parameters with obvious engineering significance. To satisfy the different requirements of course-changing and course-keeping, a fuzzy ship steering control system is developed according to three turning stages based on closed-loop gain shaping algorithm. Three subsystems are integrated by T-S fuzzy model, whose fuzzy membership functions are designed for each controller in accordance with the features of three turning stages. The proposed methodology is verified by simulation tests on nonlinear model of a multifunctional navy ship considering wind disturbances. Simulation results indicate that the system has swifter responding, less overshoot and better course-keeping performance.Considering the frequency-domain characteristic of ship roll motion response to wave disturbance is high and band-pass, a new band interference suppression method is proposed based on the closed loop gain shaping algorithm. The controllers designed by proposed method are concise and decoupling with several obvious physical significance parameters. Because the ship roll motion is an unstable non-minimum phase system, a mirror-injection method is used to design controller. Both the RRS controller and course-keeping controller are designed on a nonlinear naval ship model. Simulation tests are done in the condition of wave disturbance with different periods. The results show the controllers achieve good disturbance rejection in the ship’s natural roll frequency band, and stable performance of course keeping and changing.In order to eliminate the blight of the backlash nonlinearity in the rudder servo system to course-keeping for ships, the hybrid control strategy of the first and second order closed-loop gain shaping algorithms is used to make the best of its fine dynamic performance of the second order closed-loop gain shaping algorithm and the static error banishing of the first order closed-loop gain shaping algorithm. The simulation experiments are carried out on a nonlinear naval ship model. The simulation results show that the controller can eliminate the blight of the backlash nonlinearity under heavy sea state and reduce the peak overshoot with satisfactory robustness.In order to solve the uncertainty in ship motion control and optimize the robust course controllers, the parameters of a second order course controller based on closed-loop gain shaping algorithm is divided into model parameters and the disturbance bandwidth parameter. For the ship model perturbation, the parameters are adaptively adjusted when the responding ship model is identified online by genetic algorithm. Moreover, the bandwidth parameter is optimized to the minimize course keeping error and steering gear consumption under the influence of the sea state. The simulation tests indicate that the control system achieves higher course keeping precision, shorter setting time, smaller overshoot and better adaptivity to model perturbation and disturbance, which shows obvious significance of engineering.A concise robust collaborative optimization control system of track-keeping autopilot with RRS is brought up to solve the multi-objective optimization problem in it. Firstly, the linear responding equivalent model is advanced to replace the nonlinear ship model in controller design. The track-keeping and RRD controllers are designed based on closed loop gain shaping algorithm, and the parameters of above controller are collaboratively optimized by fast and elitist non-dominated sorting genetic algorithm (NSGA-II) with3objectives, such as the track-keeping accuracy, rudder roll reduction rate and the energy consumption of steering gear. Finally, simulation tests are done on non-linear model of a naval carrier. The results indicate that the Pareto optimal set can represent the conditionality among the objectives. And the simulations based on the optimum solution achieve better track-keeping accuracy and roll reduction rate than empirical parameters.