Research On Control Methods For Dynamic Positioning Ship With Constraints

Dynamic positioning control is the key technology to realize high-precision ocean operations for deep ocean and open ocean operations.The constrained problems caused by the limitation of physical characteristics always leads to the decrease of control performance which is the core problem affecting the high performance ocean operation and also the bottleneck restricting the implementation of the dynamic positioning ship control approaches in practical engineering.This research will consider the problems of time delay,input saturation,velocities unmeasured and dynamic uncertainties caused by the limitation of physical characteristics in dynamic positioning ship.Under these constraints,a series of controllers are designed to realize the high performance positioning or trajectory tracking control of the dynamic positioning ship.The main research work is as follows.(1)For the time delay constraint problem caused by the limitation of actuators response speed,a finite time H? positioning controller with transients is proposed by the means of Lyapunov-Krasovskii theory for dynamic positioning ship considering the constraint of state in the pre-set time interval.This controller can make the dynamic positioning ship position near the desired point in the pre-set time interval under the influence of time delay constraint and the state of ship are constrained by pre-set boundedness.Based on the finite time boundedness,the state constraint of dynamic positioning ship is guaranteed in pre-set time interval including the initial time and it makes up the limitation that the controller based on Lyapunov theory cannot quantitatively analyze the state of ship in the pre-set time including the initial time.Then,an H? performance with initial state is introduced to overcome the zero initial condition of traditional H? control while the vibration of ship can be reduced in the pre-set time.In addition,the finite time stability is studied and an improved analysis method is proposed which can reduces the conservatism of traditional method.(2)For the input saturation problem caused by the constraint of power provided by actuators,a disturbance observer-based robust adaptive multistage anticipatory anti-windup control is developed by the means of considering the mismatched disturbance.This controller can make the dynamic positioning ship track the desired trajectory accurately under the influence of input saturation constraint.The process of transformation from six-DOF dynamic positioning ship to the three-DOF one is researched and the mismatched caused by the transformation is considered firstly.This mismatched disturbance is handle by disturbance observer technology and the tracking accuracy is improved.Then,a multistage anticipatory anti-windup control is proposed integrating two anti-windup compensators with different performance to deal with the input saturation constraint.This method overcomes the overshoot caused by input saturation and futher reduce the overshoot of ship compared with the traditional anti-windup control.In addition,the dynamic surface technology is applied to reduce the computational burden.(3)For the unmeasured velocities constraint caused by the constraint of sensors in accuracy and cost,a finite time convergence state observer-based adaptive sliding mode output feedback controller is developed considering the constraints of input saturation and the rate of convergence,simultaneously.This controller can make dynamic positioning ship track to the desired trajectory rapidly under the influence of unmeasured velocities constraint and input saturation constraint.A continue finite time state observer is proposed to estimate the unmeasured velocities in finite time while the rate of convergence are improved compared with the traditional state observer and the chattering is avoided compared with the sliding mode state observer.Then,an auxiliary system is constructed to handle with the input saturation and the signal generated by the auxiliary system is introduced to design a terminal sliding mode surface while the fast convergence is achieved.The adaptive sliding mode output feedback controller is proposed based on the terminal sliding mode surface designed and the shock of ship is reduced while the rate of convergence is improved.In addition,the nominal adaptive sliding mode output feedback controller is continues which means the chattering is avoided and the tracking accuracy can be gurarnteed while the input saturation isn'tactivated.(4)For the dynamic uncertainties caused by the constraint of information about ship's physical characteristics,a minimal-parameters-learning-based robust adaptive neural network controller is proposed for dynamic positioning ship while the time delay and unknown input saturation are taken into consideration,simultaneously.This controller can make the dynamic positioning ship track along the pre-set trajectory to the desired point under multiple constraints including time delay,unknown input saturation and uncertainties.The minimal-parameters-learning technology is used to reduce the computational burden of traditional neural network control while the "explosion of complexity" and "curse of dimensionality" are avoided which will make it easy to apply the controller in practical engineering.Then,the influence caused by unknown input saturation,unknown time-varying external disturbance and the approximate error caused by neural networks are compensated by robust adaptive technology,simultaneously.It can reduce the shiock caused by unknown input saturation and the error caused by neural network is reduced,too.