Abstract:Concentrating Solar Power Tower (CSPT) plants rely on heliostat fields to focus sunlight onto a central receiver. Although simple aiming strategies, such as directing all heliostats to the receivers equator, can maximize energy collection, they often result in uneven flux distributions that lead to hotspots, thermal stresses, and reduced receiver lifetimes. This paper presents a novel, data-driven approach that integrates constraint learning, neural network-based surrogates, and mathematical optimization to overcome these challenges. The methodology learns complex heliostat-to-receiver flux interactions from simulation data, constructing a surrogate model that is embedded into a tractable optimization framework. By maximizing a tailored quality score that balances energy collection and flux uniformity, the approach yields smoothly distributed flux profiles and mitigates excessive thermal peaks. An iterative refinement process, guided by the trust region and progressive data sampling, ensures the surrogate model improves the obtained solution by exploring new spaces during the iterations. Results from a real CSPT case study demonstrate that the proposed approach surpasses conventional heuristic methods, offering flatter flux distributions and safer thermal conditions without a substantial loss in overall energy capture.

What problem does this paper attempt to address?

This paper attempts to solve the problem of optimizing the aiming strategy in the heliostat field of Concentrating Solar Power Tower (CSPT) plants. Specifically, the paper aims to overcome the problem of non - uniform flux distribution caused by existing simple aiming strategies (such as aiming all heliostats at the equator of the receiver). These problems include hot spots, thermal stress, and shortened receiver life. ### Main problems and goals of the paper 1. **Hot spots and thermal stress**: Although the traditional single - point aiming strategy can maximize energy collection, it will lead to non - uniform flux distribution on the receiver surface, generate hot spots, and then cause thermal stress, affecting the service life of the receiver. 2. **Flux distribution optimization**: A new method is needed to optimize the aiming strategy of heliostats to ensure a more uniform flux distribution on the receiver surface, thereby reducing thermal stress and prolonging the life of the receiver. 3. **Real - time performance and computational efficiency**: The existing rule - based multi - point aiming strategies or traditional optimization algorithms have the problem of high computational cost in real - time applications and are difficult to implement efficiently. ### Solutions The paper proposes a data - driven method, which combines Constraint Learning, neural network - based surrogates and mathematical optimization techniques to solve the above problems. The specific steps are as follows: 1. **Constructing a surrogate model**: Learn the complex flux interaction relationship between heliostats and receivers through simulation data and construct a neural network surrogate model. This model can be embedded in a solvable optimization framework. 2. **Optimizing flux distribution**: Define a quality score that comprehensively considers both energy collection and flux uniformity. By maximizing this score, the optimized aiming strategy can produce a smoother flux distribution and relieve excessive thermal peaks. 3. **Iterative improvement**: Adopt an iterative refinement process of trust region and progressive data sampling to ensure that the surrogate model can explore new solution spaces in each iteration, thereby gradually improving the quality of the solution. ### Main contributions - **Apply constraint learning to CSPT aiming optimization for the first time**, embed the complex data - driven heliostat - receiver flux interaction representation into a mathematical framework. - **Propose a flexible and continuous aiming strategy**, which can produce a flatter and more uniform flux distribution compared with the standard k - factor scanning heuristic method. - **Systematically balance maximizing solar energy collection and relieving thermal peaks**, reduce the risk of hot spots and ensure the long - term stable operation of the receiver. - **Combine iterative trust region and progressive sampling**, ensure the accuracy of the neural network surrogate model in the relevant areas of the design space, and thus improve the solution in each iteration. Through these innovations, the paper provides a new optimization method for the efficient and safe operation of CSPT plants and has broad application prospects.

Leveraging Neural Networks to Optimize Heliostat Field Aiming Strategies in Concentrating Solar Power Tower Plants

Heliostat Field Aiming Strategy Optimization with Post-Installation Calibration

Real-time Heliostat Field Aiming Strategy Optimization Based on Reinforcement Learning

Post-Installation Heliostat Field Aiming Strategy Optimization Based on Heliostat Flux Distribution Measurements

The Optimization Of Heliostat Canting In A Solar Power Tower Plant

An Optimal Scheduling of the Heliostat Field in A Solar Tower Power Plant

Performance Analysis of a Solar Power Tower Plant Integrated with Trough Collectors

Automatic heliostat learning for in situ concentrating solar power plant metrology with differentiable ray tracing

Optimizing the concentration ratio of multi-faceted focusing heliostats

Multi-objective Robust Optimization of a Solar Power Tower Plant under Uncertainty

Optical-thermal–mechanical comprehensive performance of the concentrating and collecting subsystem for the next-generation solar power tower based on heliostat field layouts optimization

Fuzzy Logic Energy Management Strategy of a Multiple Latent Heat Thermal Storage in a Small-Scale Concentrated Solar Power Plant

Towards high efficiency heliostat fields

Heliostat Field Layout via Niching and Elite Competition Swarm Optimization

Optimizing dispatch for a concentrated solar power tower

Heliostat layout optimization for load-following solar tower plants

Multi-objective Optimizations of Solar Receiver Based on Deep Learning Strategy in Different Application Scenarios

A Novel Sun-Tracking and Target-Aiming Method to Improve the Concentration Efficiency of Solar Central Receiver Systems

Inverse Deep Learning Ray Tracing for Heliostat Surface Prediction

Heliostat field optimization: A new computationally efficient model and biomimetic layout

Campo: Generation of regular heliostat fields