A Study on Load Position Control System Design for Developing Crane Operating Performance

Abstract

Crane systems provide indispensable lifting capabilities in a wide range of industrial fields. Among many different tasks of a crane system, the position control task for the load in the presence of unpredictable external disturbances is a crucial task. Furthermore, suppressing load vibration is also immensely critical to satisfy rigorous safeties and efficiency requirements. Therefore, the issue of improving and developing load position control performance is a very important fact. A lot of knowledge and experiments are demanded to accomplish load position control to avoid wrong operations or accidental collisions. The goal of this dissertation is to design a control system that assists the crane operator to lift the load in the vertical direction without any oscillations. In the first chapter of this dissertation, the related studies are reviewed and compared. The main difficulties associated with the load position control can be outlined as follows: • Many external factors affecting the crane system. • The nonlinear characteristics of the load position in the crane system. • The variations of the rope parameters owing to the change in the rope length, the load weight, etc. • The complication in the programming process owing to the high order of control system. • The extremely high operating safety requirements of a crane system. Based on these difficulties, two approaches are proposed to design the load position control system for developing crane operating performance by using nonlinear control theory. In the first approach, a pilot crane model equipped with only one actuator system is used to control the load position under the effects of disturbances and variations of the system parameters. The nonlinear control technique is employed to design a position controller for the load in the vertical direction. The following steps are completed to fulfill the objective of the first approach. Firstly, the motion dynamics of the load are carefully considered. Therein, the load position is compensated for by the winch actuator control. Moreover, The rope parameters in the mathematical model, which are nonlinear and strongly dependent on the change in the rope length, are estimated and calculated with time-varying. Then, the input-output feedback linearization and decoupling method are applied for coping with the hard system nonlinearity and obtaining good control performances. The simulation results reveal that the proposed control method could significantly decrease the heave displacement and accurately maintain the position of the load. Secondly, to implement in real object, a pilot crane model is built and fixed on a land site with several interconnected devices such as DC motors, encoders, and sensors. The basic configuration of system hardware for control is the acquisition card National Instruments – PCI 6229, and Lab-View programming is developed. It is worth noting that that the rope’s physical parameters used in the control system are difficult to be determined by using direct measurement methods unless these parameters are provided by the manufacturers. Therefore, a real spring is purposely inserted between the end of the rope and the load to represent the elasticity of the rope caused by the movement of the load. Then, the primary actuator model is described and estimated as a linear transfer function by implementing experiments and using the Matlab Identification Toolbox. After that, other experiments are conducted to identify the spring stiffness of the inserted spring. In this part, the controller is developed on the assumption that all needed states of the system are measured by using sensors. Next, the controller based on the input-output feedback linearization control technique is proposed to build up the fully load position control system which can handle the effect of the spring stiffness of the inserted spring and track the load target trajectory input. An experimental comparison is made between the proposed controller and existing PID control method. The experimental results indicate that the proposed strategy works well and applicable in the real field. Thirdly, in some real cases, it is impossible to accurately measure the motions of the load owing to the irregular types of the load and other environmental constraints. This means that the load position data collected by the distance sensor cannot be used in the control system to calculate the control signals. Therefore, to obtain the necessary data for the control system, the incremental encoder and the load cell are installed. Then, as well known in nonlinear control theory, the input-output feedback linearization theory is chosen to build up a full control system to get the highest performance, such as controlling load position accurately and suppressing vertical vibration of load efficiently. Furthermore, to compare with the proposed nonlinear controller, the observer-based nonlinear control technique is employed to control the position of the load by using only data from the incremental encoder. In experiments, three kinds of the control technique are conducted: the PID control technique, the proposed nonlinear control technique, and the observer-based nonlinear control technique. In the second approach, a new pilot offshore crane model, which is equipped with a sub-actuator system that copes with high-frequency residual vibrations of the load, is built. Based on this model, a position control structure for the load is proposed by using two parallel controllers for controlling two actuator systems separately: the load position controller and the anti-vibration controller. The objective of the anti-vibration controller is to reduce the residual load vibration when the load position is approaching in a specified range. The objective of the load position controller is to modify the load position following a load position reference. The Sliding Mode (SM) control technique is used to design the load position controller, where the PID control technique and the Feedback Linearization (FL) control technique are used to design the anti-vibration controller. Then, in the experiments, two different types of the reference trajectory are used: the step type and the ramp type. The effectiveness of proposed controllers is shown in experimental results.