Elman-PSO Based Predictive Control Model For Slurry Shield Tunneling In Metro Construction

The slurry shield is an important branch of the modern shield method, it is widelyapplied around the world, especially underwater tunnel in soft soil in the river and sea,slurry shield method become the most commonly used methods. The control of pressurein air chamber in slurry shield is an important link in the process of tunneling across theriver, once the control is improper, it will cause the working face collapse, the river flowbackward and a series of safety problems.Aiming at the defects of the traditional manual control, and considering thedynamic and time-varying characteristics of shield parameters, we use the neutral networkwith nonlinear mapping ability and predictive control with rolling optimization andfeedback correction characteristics to build a predictive control model. Throughestablishing the connection of the air chamber pressure and thrust, advancing speed, cutterspeed, liquid level, admission rate, and the pulp rate, to study change rules of the airchamber pressure with time, then build an intelligent predictive control system of slurryshield parameter which contains a predictor and controller. By means of controlling therelated key relevant parameters, the air chamber pressure can be predicted, we canachieve the safety control purpose in tunneling process across river.In order to improve dynamic identification and optimization ability of thepredictive control system, this paper uses Elman neural network to model the predictorand controller, and bond with the particle swarm optimization (PSO) algorithm, whichformed the Elman-PSO intelligent coupling algorithm. Then on the background ofWuhan Yangtze river tunnel project, we select the air chamber pressure, thrust, advancingspeed, penetration parameters as data sample, build an intelligent predictive controlsystem based on Elman-PSO, and with BP neural network’s results as a contrast. Theanalysis and results contrast of the engineering case show that Elman-PSO algorithm hasthe good prediction accuracy in dynamic identification and online prediction of the slurryshield parameters.The research proves that the intelligent algorithm does well in solving the nonlinearand uncertainty of complex engineering problems in civil engineering, which has a highengineering value.