Method and system for testing heat efficiency of tower type solar condensation heat absorber

Abstract

The invention relates to the technical field of solar condensation and heat collection, in particular to a method and a system for testing the heat efficiency of a tower type solar condensation heat absorber. The method comprises the following steps: a benchmark test step: forming a data set in which environment parameters and operation control data of the heat absorber correspond to heat efficiency; a machine learning step: constructing a neural network model, and learning a mapping relation from environment parameters and operation control data to thermal efficiency based on the data set training model; and a real-time analysis step: inputting the data acquired in real time into the model, and directly outputting the real-time thermal efficiency. By integrating the machine learning model, the problem that a traditional test method is complex in control logic and unstable in test due to dependence on meteorological conditions is solved, and the test speed and stability are improved.

Description

Technical Field

[0001] This invention relates to the field of solar concentrating heat collection technology, specifically to a method and system for testing the thermal efficiency of a tower-type solar concentrating absorber. Background Technology

[0002] Tower solar concentrators are a type of solar energy collection technology that uses heliostats to reflect sunlight onto the surface of a central tower receiver. The receiver, its core component, heats the working fluid to a medium-to-high temperature by receiving the concentrated solar radiation. The generated heat can be used for power generation or industrial heating. The thermal efficiency of the receiver is a key indicator of solar energy conversion performance and directly affects the energy utilization rate of the entire system. Therefore, accurate and stable testing of the receiver's thermal efficiency is crucial.

[0003] Currently, the thermal efficiency testing of tower solar concentrators typically relies on direct measurement methods. This involves collecting operational control data (such as flow rate, temperature, and pressure) of the working fluid at the receiver's inlet and outlet, along with environmental parameters, and then calculating the efficiency using thermodynamic formulas. These methods require specific operating conditions; for example, adjusting the heliostat's focusing strategy and the working fluid flow rate to achieve set values ​​at the receiver's outlet, and then calculating the efficiency based on data from steady-state conditions. However, in practical applications, this testing method is constrained by several factors: firstly, the testing process requires precise control of the heliostat's focusing point and the working fluid flow rate, making the control logic quite complex; secondly, efficiency calculations are highly dependent on the stability of meteorological conditions. For instance, fluctuations in direct normal radiation (DNI) directly affect the receiver's energy input, leading to poor repeatability and large errors in the test results. Furthermore, to minimize environmental interference, testing often needs to be conducted under ideal weather conditions, limiting the frequency and applicability of the tests. Summary of the Invention

[0004] To address the above problems, this invention provides a method and system for testing the thermal efficiency of a tower-type solar concentrator.

[0005] The first aspect of this invention provides a method for testing the thermal efficiency of a tower-type solar concentrator, comprising the following steps: Benchmarking procedures are used to generate a dataset of environmental parameters and operational control data for the receiver, corresponding to its thermal efficiency. The machine learning steps involve building a neural network model, training and updating the model based on the dataset obtained from the benchmarking steps, and enabling the neural network model to learn the mapping relationship between environmental parameters and operational control data to thermal efficiency. The real-time analysis step inputs the real-time collected environmental parameters and operation control data into the neural network model built by the machine learning step, and directly outputs the real-time thermal efficiency of the receiver.

[0006] By integrating machine learning models, this invention can output thermal efficiency quickly and stably, reducing reliance on complex control logic and meteorological conditions, and improving testing efficiency and reliability.

[0007] Optionally, the benchmarking process includes the following sub-steps: The mirror field control step involves focusing the heliostat onto the surface of the receiver and adjusting the coordinates and number of the heliostats to bring the receiver's outlet parameters to the set values. The data acquisition steps involve setting up operation control data measurement points on the receiver and setting up environmental parameter measurement points in the operating environment of the receiver, and collecting the operation control data and environmental parameters of the receiver after completing the mirror field control steps. The data processing step involves cleaning and filtering the operational control data and environmental parameters obtained from the data acquisition step, based on the requirements for the thermal efficiency test of the absorber. The data analysis step uses the operational control data and environmental parameters obtained from the data processing step to calculate the thermal efficiency of the absorber.

[0008] The above sub-steps improve the accuracy and completeness of benchmark data, providing a high-quality training foundation for machine learning models.

[0009] Optionally, the thermal efficiency of the receiver can be tested under both cloudy and clear weather conditions. The selection criteria for operational control data and environmental parameters in the data processing steps are as follows: the operational control data and environmental parameters obtained under both weather conditions are taken from the same time period on the test day; and the date interval between the two weather conditions is less than 5 days. By selecting specific weather conditions and controlling the time interval, the impact of receiver efficiency degradation and meteorological changes on the test results is reduced.

[0010] Optionally, the thermal efficiency test of the receiver is conducted between 10:30 and 13:30 true solar time at the location of the receiver; the ratio of direct normal radiation under the two weather conditions is between 0.4 and 0.8. This time window and radiation ratio condition help improve test accuracy and reduce errors.

[0011] Optionally, during the heliostat field control process, the pointing strategies of the heliostats are consistent, and the number of heliostats in the sun-tracking state is equal. This eliminates the interference of the heliostat field control strategy on the receiver thermal efficiency, achieving test consistency.

[0012] Optionally, the environmental parameters include at least direct normal radiation, ambient temperature, light intensity, and wind speed.

[0013] Optionally, the operation control data are the readings of operation control data measuring points, which include at least a flow sensor, a temperature sensor, and a pressure sensor, and the inlet and outlet of the absorber are equipped with flow regulating valves.

[0014] Optionally, during the data acquisition step, one set of environmental parameters and corresponding operational control data is recorded every minute, with a minimum of 10 sets recorded. By setting a reasonable data acquisition frequency and quantity, the sufficiency of the dataset is ensured, thereby improving the model training effect.

[0015] A second aspect of this invention provides a tower-type solar concentrating receiver thermal efficiency testing system, applying any of the aforementioned tower-type solar concentrating receiver thermal efficiency testing methods. The system includes: a receiver with an inlet and an outlet; a heliostat field to concentrate sunlight onto the receiver surface; environmental parameter measuring points located in the receiver's operating environment to monitor environmental parameters at the receiver's location; operation control data measuring points located at the receiver's inlet and outlet to monitor and collect the receiver's operation control data; a data acquisition unit to collect environmental parameters and operation control data; a data processing device to clean, filter, and analyze data from the data acquisition unit, and to perform machine learning and neural network model applications; and an operation control unit communicatively connected to the receiver and the heliostat field. This system integrates data acquisition and processing functions, supporting efficient and continuous efficiency testing.

[0016] Optionally, environmental parameter measurement points can be weather stations. Attached Figure Description

[0017] Figure 1 A flowchart of a method for testing the thermal efficiency of a tower-type solar concentrator provided in the first embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of the structure of the tower solar concentrator thermal efficiency testing system provided in the second embodiment of the present invention.

[0019] Figure reference numerals: 100-Tower solar concentrator thermal efficiency testing system, 1-Absorber, 2-Heliostat field, 21-Heliostat, 3-Inlet flow sensor, 4-Inlet temperature sensor, 5-Inlet pressure sensor, 6-Outlet flow sensor, 7-Outlet temperature sensor, 8-Outlet pressure sensor, 9-Inlet flow regulating valve, 10-Outlet flow regulating valve, 11-Data acquisition unit, 12-Data processing equipment, 13-Weather station, 14-Heliostat field control unit. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] <First Implementation Method> This embodiment provides a method for testing the thermal efficiency of a tower-type solar concentrator.

[0022] like Figure 1 As shown, the method includes a benchmarking step S1, a machine learning step S2, and a real-time analysis step S3. The benchmarking step S1 is used to generate a high-quality dataset, the machine learning step S2 constructs and trains a neural network model, and the real-time analysis step S3 applies the model to quickly output thermal efficiency.

[0023] The purpose of benchmark test step S1 is to collect a dataset of environmental parameters, operational control data, and thermal efficiency corresponding to receiver 1. Environmental parameters include direct normal radiation (DNI), ambient temperature, light intensity, and wind speed, which are monitored in real time through environmental parameter measuring points, such as weather station 13. Operational control data includes the working fluid flow rate, temperature, and pressure at the inlet and outlet of receiver 1, which are collected through operational control data measuring points, such as flow sensors, temperature sensors, and pressure sensors.

[0024] Furthermore, the benchmark testing step S1 includes four sub-steps: mirror field control step S11, data acquisition step S12, data processing step S13, and data analysis step S14, as detailed below: Heliostat field control step S11: The purpose of this step is to focus the heliostat 21 onto the surface of the receiver 1 and adjust the coordinates and number of the heliostat 21 so that the outlet parameters (such as temperature or pressure) of the receiver 1 reach the set values. This step controls the heliostat 21 through the heliostat field control unit 14 to stably focus sunlight onto the receiver 1. To eliminate the interference of the heliostat field 2 on thermal efficiency, the pointing strategy of the heliostat 21 should be consistent, and the number of heliostats 21 in the sun-tracking state should be equal. For example, in testing, the number of heliostats 21 can be dynamically adjusted according to the DNI value, but the consistency of the control strategy should still be maintained. It should be noted that, in addition to coordinate and number adjustment, other alternative solutions can be used for the focusing method of the heliostat 21, such as an automatic calibration system based on optical feedback.

[0025] Data acquisition step S12: In this embodiment, operation control data measuring points are set on the receiver 1. For example, an inlet flow sensor 3, an inlet temperature sensor 4, and an inlet pressure sensor 5 are installed at the inlet, and an outlet flow sensor 6, an outlet temperature sensor 7, and an outlet pressure sensor 8 are installed at the outlet. Simultaneously, environmental parameter measuring points are set in the operating environment, such as a weather station 13. The data acquisition unit 11 is responsible for collecting this data. The acquisition frequency is one set of parameters recorded every 2 minutes, with no fewer than 6 sets, to accumulate sufficient candidate data for subsequent steps. Preferably, the acquisition frequency can be further increased, such as recording one set of parameters every 1 minute, with no fewer than 10 sets. During data acquisition step S12, an inlet flow regulating valve 9 and an outlet flow regulating valve 10 are respectively installed at the inlet and outlet of the receiver 1 to regulate the working fluid flow rate, so that the outlet parameters, such as the outlet temperature, are stabilized at the set value. For example, tests are conducted under both cloudy and clear weather conditions, and the two regulating valves maintain consistent outlet parameters.

[0026] Data processing step S13: This step cleans and filters the collected data to determine if it meets the requirements for thermal efficiency testing. Filtering criteria include: test data must be taken under two weather conditions: a first weather condition (e.g., thin cloud cover) and a second weather condition (e.g., clear skies), with the data taken from the same time period of the same day (e.g., true solar time 10:30-13:30), and the date interval less than 5 days; the DNI ratio under the two weather conditions should be between 0.4 and 0.8. These filtering criteria help reduce the efficiency decay of receiver 1 and the impact of weather changes. It should be noted that data processing step S13 may also involve outlier removal and missing data filling to ensure data quality.

[0027] Data analysis step S14: Calculate the thermal efficiency of receiver 1 using the processed data. The calculation is based on thermodynamic formulas; for example, the power of receiver 1 is calculated using the working fluid mass flow rate and enthalpy difference, while the thermal efficiency is derived by combining the DNI ratio and the surface absorptivity of receiver 1. The specific formula is as follows: (1) (2) (3) In the formula, Indicates the power of receiver 1; This indicates the mass flow rate of the working fluid flowing in receiver 1; The enthalpy values ​​at the outlet and inlet of the working fluid in absorber 1 are indicated by the readings from the inlet temperature sensor 4, inlet pressure sensor 5, outlet temperature sensor 7, and outlet pressure sensor 8, and can be obtained from a table. These represent the receiver efficiency under the first weather condition (thin cloud cover) and the second weather condition (clear skies), respectively. This represents the ratio of direct normal radiation under the first weather condition (thin cloud cover) to the second weather condition (clear skies). The absorptivity of the surface of absorber 1 can be approximated as the absorptivity of the coating on the surface of absorber 1.

[0028] Specifically, a complete benchmark test step S1 process is as follows: At the start of the test, environmental parameters, such as direct normal radiation (DNI), ambient temperature, and wind speed, are monitored in real time by weather station 13. When the parameters indicate that receiver 1 has reached the set operating conditions (e.g., the outlet temperature setting is 400°C), the heliostat field control unit 14 calculates and adjusts the coordinates and number of heliostats 21 according to the outlet parameter setting of receiver 1. For example, when the DNI value is greater than 700 W / m², the control unit activates the sun-tracking mode, focusing all heliostats 21 on the surface of receiver 1 and maintaining a consistent pointing strategy. The number of sun-tracking heliostats 21 is fixed at 100 to ensure that the light spot stably covers receiver 1 for about 10 minutes, until the reading of the outlet temperature sensor 7 of receiver 1 stabilizes within the set value ±5°C.

[0029] While awaiting thin cloud cover, the flow rate of the working fluid (e.g., molten salt) is fine-tuned via the inlet flow regulating valve 9 to maintain the outlet temperature at the set value. Once the system is running stably, the data acquisition unit 11 starts, recording a set of parameters every minute for at least 10 minutes. The collected data includes readings from the inlet flow sensor 3, inlet temperature sensor 4, inlet pressure sensor 5, outlet temperature sensor 7, outlet pressure sensor 8, and outlet flow sensor 6, as well as data from the weather station 13, including DNI, ambient temperature, light intensity, and wind speed. Each set of data includes a timestamp to ensure synchronization. For example, data collection begins at 11:00 true solar time, recording 10 sets of data. Subsequently, under clear weather conditions with intervals less than 5 days, the same operation is repeated: during the same true solar time period (e.g., 11:00~11:10), the working fluid flow rate is controlled via the inlet flow regulating valve 9 to match the outlet temperature with thin cloud conditions, and then 10 sets of data are collected at the same frequency. All data is transmitted in real-time to the data processing device 12 for storage.

[0030] The collected raw data is cleaned and filtered. First, outliers (such as abrupt changes caused by sensor malfunctions) are removed. Then, filtering criteria are applied to retain only data sets under lightly cloudy and clear weather conditions within the same time period (preferably between 10:30 and 13:30 true solar time), with a date interval of less than 5 days, and a DNI ratio between the two weather conditions between 0.4 and 0.8. For example, if the DNI is 500 W / m² under lightly cloudy weather conditions and 800 W / m² under clear weather conditions, with a ratio of 0.625, then it meets the criteria. After data cleaning, missing value imputation (such as linear interpolation) can be performed to make the dataset complete.

[0031] Using the cleaned data, the thermal efficiency is calculated using the formula. First, the power of receiver 1 is calculated according to formula (1), which is solved by the working fluid mass flow rate and enthalpy difference (the outlet and inlet enthalpy values ​​are obtained from the working fluid thermodynamic table); then, based on formulas (2) and (3), combined with the DNI ratio and the surface absorptivity of receiver 1, the efficiency under thin clouds and clear weather is calculated respectively. After the calculation is completed, a corresponding dataset of environmental parameters, operation control data and thermal efficiency values ​​is formed and stored in the database to provide training test data samples for machine learning step S2.

[0032] A complete benchmark test step S1 outputs a set of environmental parameters, operation control data and thermal efficiency data for receiver 1. The datasets output by multiple benchmark test steps S1 are used for model training and testing in the machine learning step S2.

[0033] Machine learning step S2 constructs a neural network model, using the environmental parameters and operational control data obtained in benchmarking step S1 as input features, and thermal efficiency as the output target. The neural network model can employ, for example, a multilayer perceptron structure, including an input layer, hidden layers, and an output layer. It is trained using the backpropagation algorithm to learn the mapping relationship from input features to the output target. During training, the dataset can be divided into training and test sets to evaluate model accuracy. It should be noted that, besides neural networks, other machine learning models such as support vector machines or random forests can also be used as alternatives. After model training, it can be updated periodically to adapt to changes in the performance of receiver 1.

[0034] The real-time analysis step S3 inputs the real-time collected environmental parameters and operational control data into a well-trained neural network model, directly outputting the real-time thermal efficiency of receiver 1. This step avoids complex calculations and control logic, improving testing speed. For example, during system operation, the data acquisition unit 11 continuously transmits data to the model, and the model returns the thermal efficiency value within seconds.

[0035] Specifically, the complete process of training, applying, and updating a neural network model is as follows: During the model training phase, this phase relies on the high-quality dataset accumulated in benchmarking step S1. In benchmarking step S1, the system collects environmental parameters and operational control data of receiver 1 under strictly controlled conditions, and calculates the corresponding thermal efficiency value based on thermodynamic formulas, forming a series of "input-output" sample pairs. After cleaning and filtering, this data is used to construct a neural network model, for example, a multilayer perceptron structure, where input layer nodes correspond to the number of features, hidden layers handle nonlinear relationships, and the output layer generates predicted thermal efficiency values. During training, the model learns the complex mapping between input features and thermal efficiency and validates the fit to achieve high accuracy on the test set.

[0036] After training, the model enters the application phase, enabling rapid real-time prediction of thermal efficiency. In this phase, the well-trained neural network model is deployed to the data processing device 12 and integrated with the data acquisition unit 11. During operation, the system continuously collects real-time environmental parameters and operational control data at a fixed frequency. This data, after standardized preprocessing, is directly input into the model for inference. The model outputs the real-time thermal efficiency value of the receiver 1, eliminating the complex steady-state control and calculation processes of traditional methods and improving testing speed. For example, when solar radiation changes, the model can respond immediately and provide an efficiency estimate without waiting for manual intervention or complex calculations. This real-time analysis capability allows the system to adapt to dynamic operating conditions, providing operators with continuous and reliable efficiency monitoring.

[0037] It is worth noting that this implementation can also include an application data-driven update mechanism to maintain the model's long-term accuracy. Each new data generated in the real-time analysis step S3 (including input features and model prediction results) is automatically recorded and stored by the system, forming a continuously expanding database. This new data reflects the latest operating status of receiver 1, such as performance degradation or environmental changes. The system triggers model updates periodically (e.g., monthly) or based on performance thresholds (e.g., prediction error exceeding 5%), fine-tuning the model parameters using the new data through incremental learning or full retraining. For example, new data may include DNI fluctuations caused by seasonal changes; the model adapts to these new patterns through retraining, avoiding "aging." The entire process forms a closed loop: the application generates data, the data is used for updates, and the updates optimize the application, thereby continuously improving the model's stability and adaptability.

[0038] This implementation method integrates a machine learning model, using environmental parameters and operational data as input features, to output the thermal efficiency of receiver 1 in real time. It constructs a complete testing and application process from data acquisition to model self-updating, solving the problems of complex control logic and unstable testing due to dependence on meteorological conditions in traditional methods. It achieves high efficiency, stability, and adaptability in the thermal efficiency testing of tower solar concentrating receivers.

[0039] <Second Implementation Method> This embodiment provides a tower-type solar concentrator thermal efficiency testing system 100, which applies the testing method of the first embodiment.

[0040] like Figure 2 As shown, the system 100 includes a receiver 1, a heliostat field 2, environmental parameter measuring points, operation control data measuring points, a data acquisition unit 11, a data processing device 12, and an operation control unit (not shown in the figure).

[0041] The receiver 1 is located in the central tower and receives sunlight converged by the heliostat field 2. The receiver 1 includes an inlet and an outlet for the flow of a working fluid (such as molten salt or water). The working fluid is heated within the receiver 1, generating high-temperature thermal energy. The surface of the receiver 1 is coated with a high-absorptivity material to improve photothermal conversion efficiency.

[0042] The heliostat field 2 consists of multiple heliostats 21 that reflect sunlight onto the surface of the receiver 1. The coordinates and angles of the heliostats 21 are adjusted by the heliostat field control unit 14 to stabilize the focal point. The control unit automatically adjusts the state of the heliostats 21 based on environmental parameters to maintain the setpoints of the receiver 1's outlet parameters. It should be noted that the heliostat field 2 can be arranged in a planar or curved layout, but the control logic is primarily based on a sun-tracking strategy.

[0043] Environmental parameter measuring points are set up in the operating environment of receiver 1 to monitor environmental parameters such as DNI, ambient temperature, light intensity, and wind speed. These measuring points are typically integrated into weather station 13, with a data acquisition step size no larger than that of receiver 1 to ensure synchronization. Weather station 13 is located near the mirror field to minimize terrain interference.

[0044] Operational control data measuring points are set at the inlet and outlet of receiver 1 to monitor operational control data such as working fluid flow rate, temperature, and pressure. These measuring points include flow sensor 3, temperature sensor 4, pressure sensor 5, flow sensor 6, temperature sensor 7, and pressure sensor 8. For example, inlet flow sensor 3, inlet temperature sensor 4, and inlet pressure sensor 5 are installed at the inlet; outlet flow sensor 6, outlet temperature sensor 7, and outlet pressure sensor 8 are installed at the outlet. These sensors are calibrated with high precision to reduce measurement errors. Inlet flow regulating valve 9 and outlet flow regulating valve 10 are also installed at the inlet and outlet of receiver 1 to regulate the working fluid flow rate and maintain stable outlet parameters.

[0045] Data acquisition unit 11 connects to all measuring points, collecting environmental parameters and operational control data. Data acquisition unit 11 can employ, for example, analog-to-digital conversion and communication modules (such as wireless transmission) to achieve real-time data upload. The acquisition frequency is set to, for example, recording a set of data every 2 minutes.

[0046] Data processing device 12 is responsible for data cleaning, filtering, analysis, and machine learning model applications. Data processing device 12 can be configured as a server or embedded system, running algorithms including, for example, data preprocessing (such as anomaly detection), thermal efficiency calculation, and model training. The machine learning component integrates neural network models, and the training and inference processes can be based on platforms such as Python or MATLAB. It should be noted that data processing device 12 can be deployed in the cloud or at the edge, or locally, to reduce latency.

[0047] The operation control unit communicates with receiver 1 and heliostat field 2 to maintain stable outlet parameters of receiver 1 during testing. The operation control unit receives feedback from data processing device 12, inlet flow regulating valve 9, outlet flow regulating valve 10, and heliostat field control unit 14. For example, when DNI fluctuates, the operation control unit adjusts the working fluid flow rate to prevent efficiency deviations.

[0048] Specifically, the workflow of the tower solar concentrator thermal efficiency testing system 100 is as follows: Upon startup, environmental parameter measuring points and operational control data measuring points continuously collect data. The data acquisition unit 11 summarizes the data and sends it to the data processing device 12. The data processing device 12 first performs benchmark testing to form a dataset; then it trains a neural network model; finally, in real-time mode, the model directly outputs the thermal efficiency. The operational control unit optimizes system operation based on the efficiency value. The entire process is automated, reducing manual intervention.

[0049] The above are merely optional embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for testing the thermal efficiency of a tower-type solar concentrator, characterized in that, Includes the following steps: Benchmark testing steps: Generate a dataset corresponding to the environmental parameters and operational control data of the receiver and its thermal efficiency; Machine learning steps: Construct a neural network model, and train and update the neural network model based on the dataset obtained in the benchmarking steps, so that the neural network model learns the mapping relationship from the environmental parameters and the operating control data to the thermal efficiency; Real-time analysis step: The environmental parameters and operation control data collected in real time are input into the neural network model constructed by the machine learning step, and the real-time thermal efficiency of the heat absorber is directly output.

2. The method for testing the thermal efficiency of a tower-type solar concentrator as described in claim 1, characterized in that, The benchmarking process includes the following sub-steps: Heliostat control steps: Focus the heliostat on the surface of the receiver and adjust the coordinates and number of the heliostats so that the outlet parameters of the receiver reach the set values; Data acquisition steps: Set up operation control data measurement points on the receiver and set up environmental parameter measurement points in the operating environment of the receiver, and collect the operation control data and environmental parameters of the receiver after completing the mirror field control steps; Data processing steps: According to the requirements of the thermal efficiency test of the heat absorber, the operation control data and environmental parameters obtained in the data acquisition step are cleaned and filtered; Data analysis step: Using the operation control data and environmental parameters obtained in the data processing step, calculate the thermal efficiency of the absorber.

3. The method for testing the thermal efficiency of a tower-type solar concentrator as described in claim 1, characterized in that, The thermal efficiency test of the absorber was conducted under two weather conditions: thin cloud cover and clear skies. The data processing steps selected the operation control data and environmental parameters based on the following criteria: the operation control data and environmental parameters obtained under the two weather conditions were taken from the same time period on the test day; and the date interval between the two weather conditions was less than 5 days.

4. The method for testing the thermal efficiency of a tower-type solar concentrator as described in claim 3, characterized in that, The thermal efficiency test of the absorber is conducted between 10:30 and 13:30 true solar time at the location of the absorber; the ratio of direct normal radiation under the two weather conditions is between 0.4 and 0.

8.

5. The method for testing the thermal efficiency of a tower-type solar concentrator as described in claim 2, characterized in that, In the mirror field control step, the pointing strategies of the heliostats are consistent, and the number of heliostats in the sun-tracking state is equal.

6. The method for testing the thermal efficiency of a tower-type solar concentrator as described in claim 2, characterized in that, The environmental parameters include at least direct normal radiation, ambient temperature, light intensity, and wind speed.

7. The method for testing the thermal efficiency of a tower-type solar concentrator as described in claim 2, characterized in that, The operation control data are the readings of the operation control data measuring points, which include at least a flow sensor, a temperature sensor, and a pressure sensor. The inlet and outlet of the heat absorber are equipped with flow regulating valves.

8. The method for testing the thermal efficiency of a tower-type solar concentrator as described in claim 1, characterized in that, In the data acquisition step, one set of environmental parameters and corresponding operation control data are recorded every minute, and the number of recorded sets is not less than 10.

9. A system for testing the thermal efficiency of a tower-type solar concentrator, characterized in that, The method for testing the thermal efficiency of a tower solar concentrator as described in any one of claims 1-8, wherein the system comprises: The heat absorber includes an inlet and an outlet; A heliostat field concentrates sunlight onto the surface of the absorber; Environmental parameter measuring points are set in the operating environment of the heat absorber to monitor the environmental parameters at the location of the heat absorber; Operation control data measurement points are set at the inlet and outlet positions of the heat absorber to monitor and collect the operation control data of the heat absorber; The data acquisition unit collects the environmental parameters and the operation control data; The data processing equipment cleans, filters, and analyzes the data from the data acquisition unit, and completes the application process of machine learning and neural network models. The operation control unit is communicatively connected to the receiver and the heliostat field.

10. The tower solar concentrator thermal efficiency testing system as described in claim 9, characterized in that, The environmental parameter measuring points are meteorological stations.

Images