Prediction of agricultural implement hydraulic power requirements using Controller Area Network bus data

One of the important challenges in agricultural machinery research is the ability to effectively determine power requirements of a given field operation. The Controller Area Network (CAN) Bus, also known as ISOBUS, has proven to be an effective digital tool for tractor and implement data collection. This study attempted to determine implement hydraulic power requirements using a combination of existing public tractor CAN messages and minimal added sensors. The sensor signals were published on the CAN bus for ease of simultaneous sensor signal and CAN message data collection. Based upon the available CAN messages, this study attempted to measure hydraulic flow rate distributed by a tractor’s directional control valve without the incorporation of a flowmeter. For instances when a valve received its requested flow rate by the operator, the valve’s flow rate was predicted as a function of the valve’s spool position. The resulting curve of best fit had a coefficient of determination of 0.9993 and a root-mean-square-error of 0.8805 Lmin^-1 for the combination of multiple implement loads. When the valve became flow-limited, and the effective flow rate could no longer be determined by spool position, predicting the flow rate from a measured pressure drop across a minor loss in the system was investigated. From data collected on a flow-limited valve caused by reduced engine speed, a piece-wise line of best fit was found, predicting the flow rate based upon the pressure drop. Additional pressure sensors were used to determine the flow-state of the valve.