Parameter Identification Of Hydraulic Low Pressure Pipeline Transient Model Using Parallel Genetic Algorithms

Pressure transients accompanying cavitation and gas bubbles inside low pressure hydraulic pipelines are generally generated due to the sudden broken of the hydraulic pump or shutting up of valves in the pipeline during the working of a hydraulic system. The presence of cavitation and gas bubbles not only badly influences the performance of hydraulic pumps and systems but also affects the pressure transient behaviour in hydraulic pipelines. Consequently, the reasonable prediction of pressure transient pulsation accompanying cavitation and gas bubbles in low pressure hydraulic pipelines is of critical importance for analysing and designing of hydraulic pumps and pipelines. As Genetic Algorithms (GAs) have the disadvantages of low efficiency and time-consuming for complex optimization problem. In order to predict pressure transient pulsation inside hydraulic low pressure pipelines accompanying cavitation and gas bubbles and identify the parameters in the models exactly and high efficiently, a new method using parallel genetic algorithms (PGAs) in parameter identification for pressure transient modelling of hydraulic pipelines is presented in this paper.The mathematical models of pressure transient pulsation in hydraulic low pressure pipeline and friction item mathematical models are given in this paper. The dynamic mathematical models of cavitation volume and gas bubble volume are built. The identification parameters which are the initial gas bubbles volume, gas releasing time constant and gas dissolving time constant are confirmed. Based on the local network and Windows/MATLAB, the parallel calculation platform is built and the communication of MATLAB among a cluster of computers is realized. The fitness function is constructed based on the least-square error between experimental data and simulation results. The fitness function is calculated in parallel on a cluster of computers synchronously by integrating Distributed Computing Toolbox (DCT), Genetic Algorithms Direct Research (GADS) and Simulink. And then the PGAs is implemented in a cluster of computers based on MATLAB. By using PGAs, the parameter identification for the mathematical model of pressure transient pulsation inside hydraulic low pressure pipeline is carried out. The pressure transient pulsation mathematical model with identified parameters is obtained using PGAs. The experimental pressure pulsation curve and video data of gas bubbles and cavitation phenomenon are obtained by the experiment. Comparison of simulation results and experimental data shows that the method using PGAs to identify the unknown parameters of pressure transient models is reasonable and efficient.