Research On Digital Twin Monitoring Of Hydraulic Turbine Energy Acquisition Based On Simulation

The traditional way of generating electricity using fossil fuels is unsustainable,so we urgently need renewable energy to meet global power demand.Among many renewable sources,periodic tidal energy is one that can be developed.It does not pollute the environment and lowers carbon emissions.There are relatively few studies on the effect of wake on turbine performance,especially on energy collection coefficient and torque coefficient.Although there are some improvements in computer processing capacity and hydrodynamic calculation,considering the whole scale range,the direct numerical solution of the RANS equation is still difficult to achieve,especially in solving the simulation of tidal turbines in the real marine environment at tidal positions.Numerical simulation is usually reliable and accurate,but it takes CPU time.Hydraulic turbines face technical challenges in their engineering use.At present,the real-time monitoring methods of flow field mainly rely on sensors to monitor the parameters of the whole basin,but there are complex instrument layout and operation and maintenance costs.However,the equipment cannot cover large watersheds.Marine monitoring equipment is vulnerable to environmental damage,such as water short circuit,metal equipment corrosion,biological pollution and so on.The DT monitoring method based on simulation can well supplement the problems existing in the application of flow field test monitoring.Through a small amount of fixed-point monitoring data,the discrete distribution law of relevant parameters in the whole flow field can be given,and the operation cost is greatly reduced.DT based on simulation is the digital model of real system,which is essentially a simulation model that can be extracted and invoked quickly.However,there are few cases of combining DT with numerical simulation of tidal turbine,and there are few real-time rediction of tidal turbine flow field and monitoring of energy acquisition efficiency.The virtual simulation embedded in reality is constantly updated with the real physical model to provide a reasonable DT model of the real physical model.DT is increasingly associated with full lifecycle engineering based on real objects.Despite the efforts of researchers,the DT real-time monitoring method of leaf performance is still quite complex and not fully understood.Considering the influence of steam turbine performance deterioration,the on-site monitoring data may not match the design performance curve,which will impair the control precision and speed.Hence,if we can get accurate and fast prediction data of steam turbine energy collection parameters,we can use them as input for the control system to enhance equipment safety and reliability.Computer computing power has developed,DT is one of the means to realize fast performance mapping effectively,quickly and reliably.The DT method based on simulation model can solve the problem of fast prediction of VAT hydrodynamic performance and energy acquisition capacity.A large number of researchers have done fruitful research on the energy acquisition of wind turbines,but the DT method for fast prediction of hydraulic turbine energy capacity is still complicated and unclear because of waves and currents’ joint effect.The real-time monitoring and evaluation technology of DT is used to study hydraulic turbine performance.First,the dynamic performance of the designed HATT is studied by experiment and numerical simulation.The output flow field and energy collection efficiency of DT are verified by numerical simulation and experiments.Then,the standard particle swarm optimization algorithm and Cauchy discrete particle swarm optimization algorithm are applied.A series of extensive databases for numerical simulation are established to quickly predict and monitor the same type of turbine under different environmental conditions by DT.My SQL+Sqlyog database is used for fast data interpolation and extraction.Then the three-dimensional numerical model is simplified to a first-order digital model by POD method,and the simulation results are quickly loaded into the digital model for real-time data monitoring.We use the optimization algorithm with machine learning for the data that deviates largely from the comparison result curve.finally,the database is improved by DT.By comparing the time cost of different step size and different number of cores,it shows the advantage of DT.It provides an engineering reference for real-time monitoring of flow field distribution and hydrodynamic performance evaluation.In addition,the DT method of model visualization + CFD simulation + monitoring data is used to evaluate the energy efficiency of hydraulic turbine under the combined action of wave,current and blade rotation,in order to predict the energy performance of hydraulic turbine with fast prediction accuracy.This paper takes the vertical wave flow turbine designed by ourselves as the research object.Firstly,the performance of VWFT is evaluated by sea test,CFD simulation and DT method.We can validate the CFD simulation and DT method calculation results with the marine test results.On this basis,the prediction databases of torque,thrust,lateral force and energy acquisition coefficient of hydraulic turbine under different motion states are established.With the change of hydraulic turbine operation,the new performance data will continue to be stored in the database through deep learning(DL),which can be loaded into the digital model for rapid display.Through the quantification of time consumption,the accuracy and rapidity of the prediction method of steam turbine energy acquisition capacity under the combined action of wave and current are verified.