A feedforward-feedback control strategy based on artificial neural network for solar receivers

In a solar power tower plant, the stability of the receiver's outlet temperature is required for high-efficiency and safe operation. However, the dramatic variation of solar energy caused by clouds is a severe challenge for keeping the temperature steady at the outlet. To solve this problem, a feedforward-feedback control strategy based on an artificial neural network is proposed herein to cope with the fluctuation of solar energy by regulating the receiver's mass flow rate. The feedforward controller based on artificial neural network can quickly respond to the change of solar energy, while the feedback controller can make the receiver achieve an expected ultimate temperature. The performance of the proposed control strategy is comprehensively evaluated and compared with PID controller under different conditions. The results show that the proposed control strategy can significantly reduce the fluctuation of receiver's outlet temperature. For the step variation of direct normal irradiance ranging from −15 % to 15 %, the proposed control strategy can confine the temperature deviation within ± 1 °C. For the real dramatical and continuous direct normal irradiance variation, the temperature deviation is limited to 5 °C under the proposed control strategy, while it exceeds 30 °C under the PID controller only. The results provide an alternative efficient strategy for the receiver's mass flow control to keep a steady outlet temperature under the fluctuation of the solar resource.