Close-range, microwave radar for automatic control of base-cutter height and other cane harvester operations : SDRC Final report JCU019

Abstract

The harvesting of sugar cane is the first stage in the commercial milling of cane to produce sugar. It is well known that harvesting plays a major role in determining the overall efficiency of the sugar production process. In Australia, where virtually all sugar cane is harvested using mechanical harvesters, efficient operation of the harvester is paramount to good sugar production. One area of harvesting that has been identified as an impediment to improved harvester effectiveness is proper adjustment of the base-cutter height. Improper base-cutter setting during harvesting has a number of serious consequences for sugar production including reduced production, crop damage, additional harvester running costs and inefficient transport and milling of cane through the introduction of dirt.

The overall aim of this thesis was to develop a ground detection sensor based on microwave radar technology that could sense ground level in front of a working cane harvester. The eventual purpose of such a device would be to automatically control the base-cutter height on the cane harvester to the optimum level and so improve the efficiency of the current harvesting, farming and milling processes.

The measurement technique investigated was based on the use of a radio transmitter and receiver positioned on either side of the row of cane. The assumption was a receiver close to the ground experiences more blockage from the ground than a receiver well above ground level. Thus, it was believed that changes in received level with height could be used to detect changes in the ground position.

The project evolved in two main phases. Initially, work concentrated on verifying the sensing principle in the laboratory and later in the field. Testing verified the proposed measurement procedure with the following major conclusions. Firstly, for best results a radio signal of around 2-3GHz polarised horizontal to the ground was most suitable. This type of signal provided the best compromise between being insensitive to the presence of the cane while still allowing practical sized antennas to be employed. Secondly, field-testing showed that the cane does still effect the ideal sensor response with the orientation and condition (density, leaf matter, etc) of the cane having a noticeable influence on the measurements. These results suggested that a practical sensor would need to incorporate automatic compensation for the presence of cane and that some averaging would have to be applied to remove the random fluctuations from the underlying trends.

The second stage of the project involved the building of a prototype sensor and its’ testing on a working cane harvester. A new type of microwave ground detection sensor was developed. This device works by measuring the amplitude of a 2.4GHz, horizontally polarised microwave radio signal sent from one side of the cane row to the other. For this application, multiple receivers are stacked vertically to measure the full height profile instantaneously. The idea of using multiple receivers with some set well above ground level, was to compensate for the changing density of the cane. The transmitter and receiver antennas were based on rectangular patch arrays. The low profile of these patch antennas were ideal to be mounted flush to a harvesters’ crop divider walls. Dedicated transmitter and receiver electronics was built to generate and detect the microwave radio signals used by this system. A test system using a laptop IBM PC and running a Visual Basic 6 program to control the operation of the sensor and log results, was also developed.

The prototype sensor developed was trailed on an Austoft harvester over a one week period in the Burnett region. The main conclusions of these tests were that the sensor did work and that it could survive the harsh conditions experienced on the harvester.