Hierarchical set-point optimization and feedforward strategy for collector defocusing of a solar plant

Abstract

One of the main control objectives in parabolic trough solar thermal plants is to maintain the outlet temperature around an operating point. For this, a synthetic oil flow is used as the main control variable. However, another crucial system of the plant is the defocusing safety system of the collectors to prevent the oil temperature from exceeding an upper limit to prevent its degradation. This will occur, in general, when the oil flow reaches the maximum possible and is not able to regulate anymore the temperature. This mechanism is generally applied based on heuristic rules and partial or total defocus, which leads to a large number of actuator actions and temperature oscillations. In commercial plants, this defocus mechanism is applied firstly to the last collector, and as necessary, other collectors are defocused. In addition, it must be taken into account that loops' parameters will be, in general, different.In this work, a FeedForward-based strategy is proposed to control the outlet temperature of collectors 1, 2 and 3 of a solar plant using the defocus angle as the manipulated variable. It is also proposed to dynamically obtain the setpoint temperatures for the first 3 collectors through an optimization based on the concentrated parameter model. The results of the simulations are presented in different situations where the good performance of the strategy is observed. It is shown how the dynamic modification of the set-points can avoid possible energy losses on occasions where a fixed set-point of temperature is not the optimal option.

Data‐driven set‐point tuning of model‐free adaptive control

“…Therefore, besides of the feedback inner control loop, an outer set‐point tuning loop is provided to enhance the performance of the local MFAC. Both the design and the analysis of the proposed DDST‐MFAC are presented for a completely unknown nonlinear and nonaffine system with no need to know mechanistic model information. Consequently, the proposed DDST‐MFAC is a data‐driven method, which is different from traditional set‐point tuning control 20‐26 …”

“…However, most of the above set-point tuning methods [20][21][22][23][24][25][26] are model-based, relying on the explicit mechanistic model information of the systems. This is a disadvantage that may hinder their wide practical applications because it is not easy to build an accurate first-principle model or identify a model for a real plant that has complex dynamics, and large scales.…”

“…Consequently, the proposed DDST-MFAC is a data-driven method, which is different from traditional set-point tuning control. [20][21][22][23][24][25][26] The rest of the work is arranged as follows. Section 2 provides the problem formulation.…”

“…Reference 24 uses a genetic algorithm to optimize temperature set‐point signals for unoccupied rooms based on a simplified linear temperature model. A hierarchical set‐point optimization 25 is designed to obtain the reference temperature of a solar plant. Reference 26 investigates a set‐point optimizer for automotive air‐conditioning systems by utilizing a steady state model.…”