Solar energy as a widespread and renewable clean energy source is increasing. The technologies for the production of solar energy mainly include photovoltaic and solar heating methods [1]. Among these, the Solar Power Tower (SPT) system, a kind of solar thermal power generation technology, has already shown the capacity for large -scale electricity generation [2]. The advantages of SPT in terms of sustainability, scalability and economic viability position it as a promising candidate to become one of the most important clean and sustainable energy sources of the future.
An SPT system has a large selection of heliocide that pursue the sun, reflect on sunlight on the central recipient and concentrate in order to heat up salt temperature in the recipient. The melted salt then transfers the heat into a conventional heat cycle system, where it is converted into electricity. This centralized concentration method offers a high concentration relationship and an increased recipient surface temperatures, which leads to improved thermal efficiency [3]. At the moment, the field costs of heliostat are more than 40% of the total costs for the construction of an SPT [4]. Numerous studies have focused on improving the reflection capacity and the energy conversion efficiency of Heliostat fields and examining various aspects such as HELITUATT fieldlayout [5]Recipient parameter [6]HELITUATAT TEA [7]and heliocydris activity design [8]etc.
In a typical SPT system, heliostats typically consist of several rectangular facets that form the focus of heliocstatics in order to facilitate production, transport and installation [9]. However, if the facets are simply arranged in a flat configuration, factors such as sunset, optical errors and tracking deviations can cause the reflected river spot to be dispersed and possibly exceed the recipient area. This leads to burial losses and reduces the energy conversion efficiency of the heliocia [10]. This study aims to develop an efficient and robust optimization method to determine the optimal canting of facets in paraboloid helioceaats, with the aim of reducing the burial losses and improving concentration efficiency, which is essential for improving the economic viability and energy generation of energy generation of SPT systems.
Existing methods Canting methods of Heliostats can be divided into two categories: Dynamic Canting method and static canting method [9]. The dynamic canting method requires the installation of independent steering mechanisms for each facet, so that the real-time adjustment of your orientation is made possible [11]. However, the method causes considerable hardware costs and consumes a lot of energy during operation and is currently only used for experimental systems. The static canting method removes the canting angle of facets and mainly includes on-axis, off-axis and paraboloid types [9]. At the on-practice canting method, the normal vectors at the heliocstatism normal at a distance from the double focal point. In fact, the canting method is inspired by the solar concentrators by the shell [12]. In the off-practice canting method, the optimal facet orientation is selected and applied in a certain point in time (known as a canting moment) as a fixed orientation during the year. The performance of the off-practice canting depends on the choice of the canting moment. The Paraboloid canting method uses the concentrating properties of the parabolic surface and positions the facets after a specific parabolic canting, which reduces the diffusion of the reflected point [13].
Recent studies have analyzed and compared various methods for the canting of heliocitors. Scott A. Jones et al. Compare two canting methods (on-axle and off-axis) of the Lugo heliocales in Solar Two using the annual incident power intercept (AIPWI) as power metric [14]. Their results show that when choosing a suitable strategy for the canting of heliostat, the strategy outside the axis leads to a higher AIPWI than the on-practice strategy. Bonanos et al. Compare the annual average weighted position of static off-practice canting and the dynamic canting with the on-axis canting [15]. Their results showed that the static off-axis canting can reduce the point size by up to 5% and the heliocentation field by 1–2% compared to axis canting. In contrast, dynamic canting can reduce the point size by up to 40% and the heliocentation field by 20–25% compared to axis canting. Buck et al. Compared to four canting methods (on-axis, off-axis, paraboloid and target-oriented) with AIPWI as a metric [13]. Her study came to the conclusion that the on-up method performs the worst performance, while off-terms and parabolic methods have a similar performance. The paraboloid method is better than the off-practice method for heliocatal fields with large heliocitors [13].
With regard to the problem of determining the optimal canting parameters for various canting methods, a simulation program used with which the AIPWI values are calculated exhaustively for each parameter combination in order to identify the optimal identification [13]. For Paraboloid Canting, Buck et al. Use a comprehensive search method to determine the optimal canting parameters. Your results show that the optimization results for the Paraboloid canting follow a Gaussian distribution and have optimal parameters. Li et al. Optimized the hole punching points and sizes of the petals in the parabolic shell concentrator using a PSO+GA -Heuristic algorithm to adapt their bending tire and maximize the recipient's intercepation rate [16]. HU et al. Developed an optimization method that combines the particle swarm optimization (PSO) with the Direct Search (TR) of the trusted region to determine the optimal parabolic parameters and aim to minimize the annual burial loss [17]. The optimization process for a single heliocital act is considerably time -consuming. In addition, the river stain size of the paraboloid canting is significantly reduced compared to the axis canting, and in the PS10 field the paraboloid canting method is completely eliminated.
In summary, it can be said that the existing research on canting optimization suffers from the following topics:
- 1.
Although the above -mentioned optimization methods can solve the optimal canting parameters under different conditions, exhaustive or heuristic algorithms can significantly increase the computing costs, the above methods can only be carried out for individual heliostats and cannot efficiently master the optimization of ten thousand helost. The development of a more efficient and stable optimization method for heliostat -canting strategies therefore remains a challenge.
- 2.
In these studies, the optimization goals used an annual loss of burial and aipwi. These optimization goals are influenced by the recipient size and do not take into account the energy distribution on the river spot, which does not adequately represent the concentration of the river spot.
In order to solve the problems mentioned above, an optimization method for Paraboloid canting parameters from Heliostat is suggested. The method introduces innovatively differentiable rendering techniques from computer graphics into the optimization of the heliocy scanting. A differentiated Monte -Carlo radiation tracking algorithm, which is implemented on a GPU with the Taichi framework, is designed and applied to simulate the reflected point efficiently and precisely. The method aims to minimize the area the region, in which 95% of energy are concentrated on the reflected river stain [18]. Not only intuitively reflects the focus of the river's focus, but also takes into account the energy distribution in the river spot. With a certain smoothing method, the differentiation of the With regard to the paraboloid parameters, the gradient information of the objective function can be preserved. The algorithm of the gradient descent is then used to determine the optimal paraboloid parameters. Finally, the canting angles of the facets are calculated based on the paraboloid parameters, which optimizes the heliostat -canting strategy. The method shows a significant optimization efficiency in both individual helmet and heliocaral field optimization problems. In summary, the method offers an efficient solution for optimizing the heliotat canting in SPT systems. The rest of the paper is organized as follows: Section 2 defines the problem of the heliostat -canting optimization, presents the decision -making variables and the optimization goal and discusses its differentifiable processing methods. Section 3 provides the details of the framework and implementation detail of the differentiable algorithm to pursue rays. Section 4 contains experimental results, comparisons, analyzes and discussions; Section 5 completes the paper and outlines future work.