الفهرس 
Flow Control System Configuration
Predictive Control Scheme Based on Neural NetworkOnly 14 pages are availabe for public view
 |  | from | 173 |  |  |
| | | from | 173 | | |
Abstract
Energy efficiency has a prominent role in fluid power researches. One technique to improve the efficiency is through decreasing flow control losses, a unique way to do so is called Digital Hydraulic Valves (DHV), in which the output flow is the summation of smaller discrete values exhausted from parallel connected valves with common input and common output nodes. The valves configuration is based on the Pulse Code Modulation (PCM) control, although this scheme is characterized by small package size and low energy consumption, it faces challenges such as higher flow peaks more than the other coding schemes, and thus resulting in poor tracking accuracy. This dilemma manifests once a transition occurs between specific flow rates.
These problems can be solved by introducing a delay in the signal opening the following valve, subsequently giving the previous valve a sufficient time to ensure its closure and hence decreases the overlapping. The delayed signal samples feeding the feedforward Neural-Network controller is acquired through Genetic Algorithm. The controller role appears in regulating the speed of a hydraulic cylinder through controlling the Digital Flow Control Unit (DFCU) valves. The DFCU —placed in a metering-in circuit— has five on/off valves. The analysis has been executed with the aid of MATLAB Simscape Fluids.
The results indicated a dropping of 93% in energy consumption for DFCU when compared with a proportional valve. The results also show a good tracking accuracy, wherein some conditions, the error decreased by one third. On the other side, the system experiences low accuracy at higher acceleration demands and sudden loads; thus, it requires fast switching valves. However, the studied system succeeded in raising the accuracy slightly at these conditions.
In conclusion flow control via employing parallel connected on/off valves has witnessed an increase in accuracy and efficiency, however, a high switching with low response time valves are recommended.