Modeling And Optimization Of Solar-Thermal Subsystem In Solar Tower Power System

Solar tower power system is a common large-scale solar thermal power system, it concentrates sun lights to the receiver by a large number of heliostats, heating the heat transfer fluid (HTF) in the tube and generating high-temperature steam to drive turbines to generate power. This solar-based power generation form has the advantages of non-pollution and huge prospect for sustainable development. Solar tower power system mainly consists of two parts: solar-thermal subsystem and power cycle subsystem. The solar-thermal subsystem, which contains heliostat and receiver, plays a crucial role of converting solar energy to heat and directly affects the subsequent systems, have a greater impact on overall system performance. Therefore, there is realistic significance to conduct research on the concentrating simulation and the optimization strategy of the concentration and collection subsystems.In this thesis, the modeling and optimization of the solar-thermal subsystem are researched and the main achievements are as follows.(1) A light-heat energy conversion model is established by integrating the heliostat model and the receiver model. The heliostat model gets the receiver surface energy flux density based on Monte Carlo ray tracing method. The receiver model calculates the temperature of molten salt based on heat transfer mechanism of the cylindrical receiver. Parallel acceleration based on GPU (Graphics Processing Unit) is used to improve the calculation speed of simulation. The correctness of the model was verified by comparing the simulation results to the test data of actual system operation.(2) On the basis of the light energy to heat energy model, an optimization model is presented to achieve the maximization outlet molten salt temperature of the receiver under 600 ℃ and the uniform energy distribution of the receiver surface by adjusting the position of aiming points, ensuring that the overheating damage can be avoided. Sequential quadratic programming (SQP) is used to solve the optimization problem. Simulation results show that the receiver surface energy distribution uniformity is significantly improved and the optimal outlet molten salt temperature can be achieved.(3) Compound method and MBDFO (Model-Based Derivative-Free Optimization) algorithm are improved and applied separately to solve the problem of convergence difficulty and long solving time in the SQP algorithm. The simulation results contrast the two algorithms and show that both of them can solve the convergence difficulties problem, while the improved model-based DFO algorithm is more efficient than the improved compound method because the latter needs to calculate objective functions frequently.