Control of cranes with a swinging load, an investigation of pragmatic control

Control of cranes with a swinging load, an investigation of pragmatic control

In modern industrial systems, cranes are widely used for the transference of heavy loads. There are different types of cranes which are used, such as tower cranes and gantry cranes. During routine operations, it is difficult for crane operators to lift and manoeuvre heavy payloads without causing significant load swing. To ensure the safety of workers and to prevent damage on site, it is essential to suppress the swinging load during operation.

Systems used in the past to control movement and load swing have used complex control algorithms requiring massive computations. This research investigates the application of ‘pragmatic’ control techniques to crane systems which represents a new method for controlling crane movement have been applied to crane systems for the first time. A new design of a pragmatic control such as a cascaded constrained loop is proposed for crane systems which are not currently being used in such systems. A simple and robust pragmatic controller has been designed to control a crane system so that the controller could move a swinging load from one point to another in minimum time without generating a large swing.

For the purposes of this research, two physical models have been constructed, in addition to considerable simulation studies. One model is applicable for a tower crane with a hoist that travels along a jib that rotates on a vertical axis. The other is a single axis model, with one axis operating on a gantry system that has been used to investigate the finer points of the control.

This work presents a state-space control method that can be implemented on a simple microcontroller. It used computer vision to monitor the crane load position and then used that position as feedback in a control loop. It has been proposed that a camera, mounted on the hoist and looking vertically downwards, can detect a corner-reflector mounted on the hook to give the essential signals for stabilisation. The development of control techniques, in particular ‘pragmatic control’, can then be central to autonomous control or auto-stabilised manual control.

There is considerable discussion of the generation of the ‘pragmatic’ control algorithm and its simplification. Also, many hardware-based techniques are developed, including software for extracting state-space data from the camera and for reading the two-phase incremental encoders of the motors at high speed.

The short time-constants associated with the smaller laboratory models caused a problem because of the slow frame-rate of the camera, thereby requiring a more sophisticated discrete-time method was required for them. Practical testing validated the control strategy and successful and robust performance was achieved. Loads were transported and positioned successfully, with little swing evident during operation of the crane.

Mathematical models of a crane in state-space are presented. A ‘pragmatic’ controller was successfully used to control the crane in simulation and then in a laboratory. Results were evaluated and compared to published benchmarks.