Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
The method and the device for predicting the photothermal power generation power in the embodiment of the invention are directed to a tower type photothermal power generation system, and in order to better describe the method and the device for predicting the photothermal power generation power in the embodiment of the invention, the tower type photothermal power generation system is described below. Fig. 1 is a schematic structural diagram of a tower-type photo-thermal power generation system, as shown in fig. 1, sunlight emitted by the sun is reflected by a heliostat, received by a solar receiver arranged at the top of a tower, and heated by a heat transfer medium to generate superheated steam, which drives a turbine generator to generate power and input the power into a power grid, so that the absorbed solar energy is converted into electric energy.
According to the embodiment of the invention, aiming at a photo-thermal conversion link and a turbine power generation link in a tower type photo-thermal power generation system, a Support Vector Machine (SVM) numerical model is adopted for prediction, so that a predicted value of the power generation power of the photo-thermal power generation system is obtained. The SVM algorithm is a supervised learning model and related learning algorithm for analyzing data in classification and regression analysis. Given a set of training instances, each of which is labeled as belonging to one or the other of two classes, the SVM training algorithm creates a model that assigns the new instance to one of the two classes, making it a non-probabilistic binary linear classifier. The support vector machine SVM model represents instances as points in space, so that the mapping is such that the instances of the individual classes are separated by as wide an apparent interval as possible. The new instances are then mapped to the same space and the categories are predicted based on which side of the interval they fall on. Support vector machines construct hyperplanes or hyperplane sets in high-dimensional or infinite-dimensional space, which can be used for classification, regression, or other tasks.
The method and the device for predicting the photothermal power generation power according to the embodiment of the invention are described in detail below with reference to the drawings.
Example one
Fig. 2 is a schematic flow chart of an embodiment of a method for predicting photothermal power provided by the present invention. As shown in fig. 2, the method for predicting the photothermal power generation power according to the embodiment of the invention may specifically include the following steps:
s201, predicting the temperature of the heat transfer medium according to meteorological data and heliostat control parameters.
Specifically, the step provides how to accurately calculate the heat quantity of the heat transfer medium converted from the collected light energy aiming at the photo-thermal conversion link. The temperature of the heat transfer medium can be predicted according to meteorological data and heliostat control parameters.
S202, predicting the generating power of the photo-thermal generating system according to the temperature of the heat transfer medium and the flow of the heat transfer medium.
Specifically, the step provides how to accurately calculate the electric energy converted from the heat of the heat transfer medium in the power generation link of the turbine. Specifically, the generation power of the photothermal power generation system can be predicted according to the predicted temperature of the heat transfer medium and the obtained flow rate of the heat transfer medium.
It should be noted here that, in the method for predicting photothermal power generation power according to the embodiment of the present invention, the problems of time lag and heat dissipation loss in the heat transfer and storage link are ignored, that is, the temperature of the heat transfer medium output from the tower section is equal to the temperature of the heat transfer medium at the turbine end, that is, the temperature of the heat transfer medium in the embodiment of the present invention is equal to the temperature of the heat transfer medium output from the tower section and is also equal to the temperature of the heat transfer medium at the turbine end.
The method for predicting the photo-thermal power generation power, provided by the embodiment of the invention, aims at the photo-thermal conversion link, predicts the temperature of a heat transfer medium according to meteorological data and heliostat control parameters, and realizes the prediction of the photo-thermal conversion link on the heat energy converted by the photo-thermal energy. Aiming at the power generation link of the turbine, the power generation power of the photo-thermal power generation system is predicted according to the temperature of the heat transfer medium and the flow of the heat transfer medium, so that the prediction of the power generated by converting heat energy into electric energy in the power generation link of the turbine is realized. In summary, the method for predicting the photo-thermal power generation power of the embodiment of the invention can predict the power generation power of the photo-thermal power generation system more accurately.
Example two
Fig. 3 is a schematic flow chart of a method for predicting photothermal power provided by the present invention according to another embodiment. The method for predicting the photothermal power generation power according to the embodiment of the invention is a specific implementation manner of the method for predicting the photothermal power generation power according to the embodiment shown in fig. 2. As shown in fig. 3, the method for predicting the photo-thermal power generation power of the embodiment of the invention may specifically include the following steps:
step S201 in the embodiment shown in fig. 2 may specifically include the following steps: and predicting the temperature of the heat transfer medium based on the photo-thermal numerical model according to meteorological data and heliostat control parameters.
Specifically, in the photothermal conversion link, the temperature of the heat transfer medium is predicted according to meteorological data and heliostat control parameters, and specifically, the temperature can be predicted by adopting a specific calculation formula or by using a photothermal numerical model obtained by machine learning training. The photothermal numerical model may be a Support Vector Machine (SVM) numerical model.
Further, based on the photothermal numerical model, the step S201 may specifically include the following steps S301 and S302.
S301, training to obtain a photo-thermal numerical model by taking historical data of meteorological data and historical data of heliostat control parameters as input and historical data of the temperature of a heat transfer medium as output.
Specifically, the step is a training process of a photo-thermal numerical model of the photo-thermal conversion link. Historical data of meteorological data and historical data of heliostat control parameters are used as input of a photothermal numerical model of the photothermal conversion link, historical data of the temperature of a heat transfer medium is used as output of the photothermal numerical model of the photothermal conversion link, and a numerical algorithm such as a Support Vector Machine (SVM) algorithm is adopted to train and obtain the photothermal numerical model of the photothermal conversion link.
And S302, taking the predicted value of the meteorological data and the predicted value of the heliostat control parameter as input, and obtaining the predicted value of the temperature of the output heat transfer medium based on the photo-thermal numerical model.
Specifically, the step is a prediction process of a photo-thermal numerical model of the photo-thermal conversion link. And (3) taking the predicted value of meteorological data (from weather forecast) and the predicted value of heliostat control parameters (from a control system) as the input of a photo-thermal numerical model of the photo-thermal conversion link, predicting based on the trained photo-thermal numerical model of the photo-thermal conversion link, and obtaining the predicted value of the temperature of the heat transfer medium output by the photo-thermal numerical model of the photo-thermal conversion link.
The support vector machine SVM includes a Support Vector Classification (SVC) and a Support Vector Regression (SVR), both of which are linear and nonlinear problems. The invention adopts a Support Vector Regression (SVR) to model. For a given set of training samples { (x) 1 ,y 1 ),(x 2 ,y 2 ),…,(x 3 ,y 3 ) In which x is i ∈R n The data are meteorological data and historical data of heliostat control parameters, and are in a two-dimensional array format; y is i The element belongs to R, namely historical data of the temperature of the heat transfer medium, and is in a one-dimensional array format; 1, 2, 3 … …, a simple linear regression function can be expressed as:
f(x)=w*x+b(1)
in the formula (1), w is a weight coefficient, i.e., a weight, and b is a deviation. For the non-linear regression problem, the basic idea of support vector machine SVM is to transform the input space to a high dimensional space using a non-linear transformation defined by an inner product function. Finding a non-linear relationship between the input variables and the output variables in a high-dimensional space:
f(x)=w·Φ(x)+b(2)
in equation (2), Φ (x) is a nonlinear transformation from an input space to a high-dimensional space.
In the conventional SVM algorithm, this low-dimensional to high-dimensional conversion is realized by a kernel function. Commonly used kernel functions are:
linear kernel function: k (u, v) ═ u.v)
Polynomial kernel function: k (u, v) ═ r (u.v) + coef0) d
RBF kernel function: k (u, v) ═ exp (-r | u-v- 2 )
Sigmoid kernel function: k (u, v) ═ tanh (r (u-v) + coef0)
The linear kernel function is mainly used in the linear separable case, which is not suitable for the present invention, and the present invention discusses the non-linear problem. The case where a polynomial kernel can be used is a simpler nonlinear case, difficult to use for complex cases, and therefore not applicable to the present invention. The RBF kernel function can be used in various situations, is the most widely applied kernel function, has good performance and shows good performance in practical problems, so the invention can adopt the commonly used RBF kernel function to realize the conversion from low dimension to high dimension.
A Support Vector Machine (SVM) numerical model among meteorological data, heliostat control parameters and the temperature of a heat transfer medium can be established through the processes, then actual historical data are substituted into the model, and a numerical matrix of the weight w and the deviation b can be obtained through training and learning of an SVM algorithm. After the weight w and the deviation b are obtained, the predicted value of the meteorological data and the predicted value of the heliostat control parameter are substituted into the model, and the predicted value of the temperature of the heat transfer medium can be calculated. The specific operation flow can be as shown in fig. 4, and model training is performed based on the historical data set and the constructed initial model to obtain an initial prediction model. And testing the initial prediction model based on a test data set (the test data set is actually historical data, and only divides the historical data into two parts, namely a large part and a small part, wherein the large part is used for learning and training, and the small part is used for verification), determining whether the model is available, and if so, determining the initial prediction model as a final prediction model. And performing photo-thermal conversion prediction based on the prediction data set and the final prediction model.
It should be noted here that the support vector machine SVM algorithm is adopted for modeling in the present invention, and in practical application, a neural network algorithm may be used, the operation flow is similar to that of the present invention, which method is specifically selected, needs to be determined according to the specific situation of the electric field, and the two methods are used for modeling and then comparing the results, and the method with better prediction result is selected.
Further, the meteorological data may specifically include, but is not limited to, any one or more of the following: irradiance, cloud cover, temperature, humidity, wind speed, and the like.
Further, the heliostat control parameter may specifically be a heliostat angle, and the like.
Step S202 in the embodiment shown in fig. 2 may specifically include the following steps: and predicting the generated power of the photo-thermal power generation system based on the thermoelectric numerical model according to the temperature of the heat transfer medium and the flow of the heat transfer medium.
Specifically, in the power generation link of the turbine, the power generation power of the photo-thermal power generation system is predicted according to the temperature of the heat transfer medium and the flow of the heat transfer medium, and the prediction can be specifically carried out by adopting a specific calculation formula or a thermoelectric numerical model obtained by machine learning training. The thermoelectric numerical model can be a Support Vector Machine (SVM) numerical model.
Further, based on the thermoelectric numerical model, step S202 may specifically include the following steps S303 and S304.
And S303, training to obtain a thermoelectric numerical model by taking the historical data of the temperature of the heat transfer medium and the historical data of the flow of the heat transfer medium as input and the historical data of the generating power of the photo-thermal generating system as output.
Specifically, the step is a training process of a thermoelectric numerical model of the turbine power generation link. Historical data of the temperature of the heat transfer medium and historical data of the flow of the heat transfer medium are used as input of a thermoelectric numerical model of the turbine power generation link, historical data of the power generation power of the photothermal power generation system is used as output of the thermoelectric numerical model of the turbine power generation link, and the thermoelectric numerical model of the turbine power generation link is obtained through training by adopting a numerical algorithm such as a Support Vector Machine (SVM) algorithm. The training process of above-mentioned light and heat conversion ring festival support vector machine SVM numerical model can be referred to specific training process, and input and output variable is different only, and here is no longer repeated.
And S304, taking the predicted value of the temperature of the heat transfer medium and the predicted value of the flow of the heat transfer medium as input, and obtaining the predicted value of the output generated power of the photo-thermal power generation system based on the thermoelectric numerical model.
Specifically, the step is a prediction process of a thermoelectric numerical model of a turbine power generation link. And (4) taking the predicted value of the temperature of the heat transfer medium (from the predicted result of the step (S302)) and the predicted value of the flow rate of the heat transfer medium (from the control system) as the input of a thermoelectric numerical model of the turbine power generation link, predicting based on the trained thermoelectric numerical model of the turbine power generation link, and obtaining the predicted value of the power generation power of the photo-thermal power generation system output by the thermoelectric numerical model of the turbine power generation link. For a specific prediction process, reference may be made to the prediction process of the above-mentioned photothermal conversion joint support vector machine SVM numerical model, but input and output variables are different, and details are not described here.
Further, in step S303, when the thermoelectric numerical model is obtained by training, the historical data of the parameters as input may further include, but is not limited to, historical data of any one or more of the following parameters: the heat exchanger heat transfer medium side inlet temperature, the heat exchanger heat transfer medium side outlet temperature, the heat exchanger turbine side inlet temperature, the heat exchanger turbine side outlet temperature, the heat exchanger turbine side inlet pressure, the heat exchanger turbine side inlet flow and the like.
Further, in step S304, when performing prediction based on the thermoelectric numerical model, the predicted value of the parameter as input may further include, but is not limited to, the predicted value of any one or more of the following parameters: the inlet temperature of the heat transfer medium side of the heat exchanger, the outlet temperature of the heat transfer medium side of the heat exchanger, the inlet temperature of the turbine side of the heat exchanger, the outlet temperature of the turbine side of the heat exchanger, the inlet pressure of the turbine side of the heat exchanger, the inlet flow of the turbine side of the heat exchanger and the like.
Further, the heat transfer medium may specifically be molten salt or the like.
The method for predicting the photo-thermal power generation power, provided by the embodiment of the invention, aims at the photo-thermal conversion link, predicts the temperature of a heat transfer medium according to meteorological data and heliostat control parameters, and realizes the prediction of the photo-thermal conversion link on the heat energy converted by the photo-thermal energy. And aiming at the turbine power generation link, the power generation power of the photo-thermal power generation system is predicted according to the temperature of the heat transfer medium and the flow of the heat transfer medium, so that the prediction of the turbine power generation link on the electric energy converted from the heat energy is realized. In conclusion, the method for predicting the photo-thermal power generation power can accurately predict the power generation power of the photo-thermal power generation system.
EXAMPLE III
Fig. 5 is a schematic structural diagram of an embodiment of the device for predicting photothermal power provided by the present invention. The device for predicting photothermal power generation power according to the embodiment of the invention may be used to perform the method for predicting photothermal power generation power according to the first or second embodiment. As shown in fig. 5, the device for predicting photothermal power according to the embodiment of the invention may specifically include: a first prediction module 51 and a second prediction module 52.
And a first prediction module 51 for predicting the temperature of the heat transfer medium based on the meteorological data and heliostat control parameters.
The second prediction module 52 is configured to predict the generated power of the photo-thermal power generation system according to the temperature of the heat transfer medium and the flow rate of the heat transfer medium.
Further, the first prediction module 51 may be specifically configured to:
and predicting the temperature of the heat transfer medium based on the photo-thermal numerical model according to meteorological data and heliostat control parameters.
Further, the first prediction module 51 may be specifically configured to:
training to obtain a photo-thermal numerical model by taking historical data of meteorological data and historical data of heliostat control parameters as input and historical data of the temperature of a heat transfer medium as output;
and taking the predicted value of the meteorological data and the predicted value of the heliostat control parameter as input, and obtaining the predicted value of the temperature of the output heat transfer medium based on the photo-thermal numerical model.
Further, the meteorological data may specifically include, but is not limited to, any one or more of the following: irradiance, cloud cover, temperature, humidity, wind speed, and the like.
Further, the heliostat control parameter may specifically be a heliostat angle, and the like.
Further, the second prediction module 52 may be specifically configured to:
and predicting the generated power of the photo-thermal power generation system based on the thermoelectric numerical model according to the temperature of the heat transfer medium and the flow of the heat transfer medium.
Further, the second prediction module 52 may be specifically configured to:
training to obtain a thermoelectric numerical model by taking historical data of the temperature of the heat transfer medium and historical data of the flow of the heat transfer medium as inputs and historical data of the power generation power of the photo-thermal power generation system as outputs;
and obtaining a predicted value of the output generated power of the photo-thermal power generation system based on the thermoelectric numerical model by taking the predicted value of the temperature of the heat transfer medium and the predicted value of the flow rate of the heat transfer medium as input.
Further, when the second prediction module 52 trains the thermoelectric numerical model, the historical data of the parameters as input further includes historical data of any one or more of the following parameters: the heat exchanger heat transfer medium side inlet temperature, the heat exchanger heat transfer medium side outlet temperature, the heat exchanger turbine side inlet temperature, the heat exchanger turbine side outlet temperature, the heat exchanger turbine side inlet pressure and the heat exchanger turbine side inlet flow;
when the second prediction module 52 performs prediction based on the thermoelectric numerical model, the predicted values of the parameters as inputs further include predicted values of any one or more of the following parameters: the heat exchanger heat transfer medium side inlet temperature, the heat exchanger heat transfer medium side outlet temperature, the heat exchanger turbine side inlet temperature, the heat exchanger turbine side outlet temperature, the heat exchanger turbine side inlet pressure, and the heat exchanger turbine side inlet flow.
Further, the heat transfer medium may specifically be a molten salt or the like.
Specifically, the specific process of the modules in the photo-thermal power generation power prediction device according to the embodiment of the present invention to realize the functions thereof can be referred to the photo-thermal power generation power prediction method shown in the first embodiment or the second embodiment.
The device for predicting the photo-thermal power generation power, provided by the embodiment of the invention, is used for predicting the temperature of a heat transfer medium according to meteorological data and heliostat control parameters in a photo-thermal conversion link, so that the prediction of the photo-thermal conversion link on the heat energy converted into the heat energy is realized. Aiming at the power generation link of the turbine, the power generation power of the photo-thermal power generation system is predicted according to the temperature of the heat transfer medium and the flow of the heat transfer medium, so that the prediction of the power generated by converting heat energy into electric energy in the power generation link of the turbine is realized. In conclusion, the device for predicting the photo-thermal power generation power can predict the power generation power of the photo-thermal power generation system more accurately.
Those of ordinary skill in the art will understand that: all or a portion of the steps of implementing the above-described method embodiments may be performed by hardware associated with program instructions. The program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.