Search Results

In many areas, neural network technology has made a successful transition from theory to practical application, primarily due to the advances that have been made in computer technology and digital signal processing. Research at the University of Saskatchewan over the past few years has focused on applying neural network technology to fluid power systems. This paper will examine four projects that have been initiated by the authors and their graduate students which use neural networks for purposes of open loop pattern following, multiple input - multiple output control, indirect measurement of actuator displacement, and hydraulic component identification. A brief introduction to static and dynamic neural networks is given. Descriptions of the individual project objectives, the experimental implementation of neural networks to achieve these objectives, and some typical experimental results are considered.

Most fertilizer application systems are not capable of variable rate adjustments “on-the-fly”. To change the application rate, the farmer must dismount the tractor and change the gear ratio mechanically (i.e. via gears, chains, etc.). Air seeder manufacturers have come up with their own unique solutions to address this problem, usually involving electrohydraulics. At present there are older seeding units that perform adequately, but do not have the variable rate option. A retrofit is therefore very desirable for these units. In this paper, the feasibility of a simple hydraulic proportional valve and variable speed motor circuit is employed to replace the gears and chains. The unit is integrated with a microcontroller to provide compensation to the nonlinear properties of a proportional valve, and in turn provide a very accurate feedrate. In addition, direct user input from the cab of the tractor is possible, allowing on-the-go rate changes.

The potential exists for significantly reducing operating costs by minimizing missing and overlap in successive agricultural field operations, particularly in areas where the equipment is pull-behind and widths of over fifteen meters are common. This paper reports on the development of a sensor, consisting of a video camera and image processing system, to detect the location of the demarcation line between tilled-and-untilled soil and cut-and-standing crop. The development of hardware and software to achieve real-time operation under a variety of crop, soil and ambient lighting conditions is described.

The agricultural combine harvester is one of the most complex of machines used in cereal grain production. It is believed that significant increases in efficiency and productivity could be realized if several of the machine adjustments could be automatically controlled. One of the most important parameters to measure and control is the feedrate, i.e. the mass flow rate of material entering the machine. This paper reports on an investigation of a sensor, using electrical capacitance techniques, to measure feedrate.

The Nexus vehicle was designed and built for Transport Canada at the University of Saskatchewan to demonstrate that a safety compliant single passenger commuter vehicle could attain extremely low fuel consumption rates at modest highway speeds. Experimentally determined steady state fuel consumption rates of the Nexus prototype ranged from 1.6 L/100 km at 61 km/hr up to 2.8 L/100 km at 121 km/hr. Fuel consumption rates for the Society of Automotive Engineers (SAE) driving cycle tests were 4.5 L/100 km for the SAE Urban cycle and 2.0 L/100 km for the SAE Interstate 55 cycle. The efficiency of the power train was determined using a laboratory dynamometer, enabling the road test results to be compared to the results from an energy and performance simulation program. Predicted fuel economy was in good agreement with that determined experimentally. Widespread use of single passenger commuter vehicles would substantially reduce current transportation energy consumption.

This paper describes the development and preliminary field evaluation of a combine feedrate sensor. The sensor employs electrical capacitance sensing techniques in a parallel plate configuration in which one plate is located in the table (or platform) with the other being the table auger. The sensor indicates the mass flowrate (feedrate) of material conveyed by the auger to the feeder. For comparison purposes, the combine harvester was intrumented for measurements of table auger torque and feeder displacement. Conventional bagging techniques of effluent over fixed time intervals were used to establish average feedrate. The results show that the capacitance sensor provides an indication of feedrate and more linear performance over a wider range of feedrate than the other two sensing techniques.

This paper compares the characteristics of a high-precision hydrostatic actuator to that of conventional hydraulic systems using servovalves. Servovalve controlled hydraulic actuators retain their market share as they provide precision movement and offer a very high torque to mass ratio at the final actuation point. The input current/output torque relationship of a conventional hydraulic actuation system is reviewed in a robotic context. This relationship is summarized by a mathematical model that can be expressed in a generalized form. This model is used for the analysis of flow and dynamic characteristics. The design and modeling of a recently proposed high-performance hydrostatic actuation system referred to as the ElectroHydraulic Actuator (EHA) is briefly reviewed. A prototype of this actuator has been produced and has demonstrated a comparable performance to servovalve controlled conventional hydraulic systems.