Monitoring Digital Technologies in Hydraulic Systems Using CUSUM Control Charts

Abstract

Digital fluid power is a growing field which utilizes electronics and advanced controls to improve efficiencies, energy savings, and productivity in fluid power systems. Often relying on on/off high-speed switching techniques, digital hydraulics relies heavily on the performance of valves, where an error in the valve performance could lead to a major drop in the efficiency and performance of the entire system. Specifically, digital pump/motors are sensitive to valve delay and transition timing which negatively impacts their performance and condition with time. One approach to assessing the performance and efficiency of digital pump/motors is via monitoring its inlet (low) and outlet (high) pressure time-series. Real-time condition monitoring also supports preventive maintenance and provides a better understanding of the dynamics of pump/motors. For condition monitoring, Statistical Process Control (SPC) charts are often designed to detect shift changes in time-series. This paper proposes to construct two cumulative sum (CUSUM) control charts for fast real-time shift detection in the high and low pressure time-series of digital pump/motors. The proposed method will be able to actively detect common misbehaviors in the valves utilized in the digital pump/motor. The model have been successfully tested on a three-piston inline digital pump/motor, but this monitoring technique can be modified and implemented on other digital technology classes where valve performance is key in the success of the system.