Data processing method, device and electronic equipment for solar thermal power generation system

Abstract

The present application discloses a data processing method, device, and electronic device for a solar thermal power generation system, relating to the field of new energy technology. The method includes obtaining historical meteorological data and historical DNI data of a target solar thermal power generation system; determining the daily variation relationship of the lag range between the temperature data and the DNI data over time based on the historical temperature data and the historical DNI data in the historical meteorological data; correcting the historical temperature data based on the daily variation relationship to obtain target temperature data; replacing the historical temperature data in the historical meteorological data with the target temperature data, and using the processed historical meteorological data and historical DNI data as input data for model training or DNI parameter prediction. Through the above method, dynamic correction processing of historical temperature data can be achieved, the correlation between historical temperature data and DNI data can be enhanced, and the prediction accuracy of the model or the accuracy of DNI parameter prediction can be improved.

Description

Technical Field

[0001] The present application relates to the field of new energy technology, and in particular to a data processing method, device, and electronic equipment for a solar thermal power generation system. Background Art

[0002] Direct normal irradiance (DNI) is a core parameter for the energy capture efficiency of CSP systems. Its prediction accuracy directly impacts the scheduling of concentrator fields, the regulation of thermal storage systems, and grid stability. With the accelerated transformation of energy structures and the continued growth of CSP installed capacity, high-precision DNI prediction has become a key technical requirement for improving the economic efficiency of CSP plants and grid reliability.

[0003] Current DNI prediction models typically rely on machine learning or physics-driven models based on meteorological time series data (such as temperature, humidity, and aerosol concentration). However, current models significantly underdescribe the dynamic interactions between these multiple factors, particularly the nonlinear time-series correlation between temperature and DNI. Specifically, solar radiation reaching the ground is first absorbed and converted into thermal energy by the surface, which then causes the surface temperature to rise. This rise or fall in temperature lags behind the solar radiation, causing the measured temperature series to lag behind the DNI series. Current models generally ignore this lag effect and directly input the raw temperature time series data into the prediction model, resulting in a time mismatch between the input features and the target DNI. This feature lag can lead to systematic prediction bias, reducing the model's ability to respond to sudden DNI events, and thus impacting the scheduling of concentrator fields, the regulation of thermal storage systems, and grid stability.

[0004] Currently, no effective solution has been proposed to the problem of low DNI prediction accuracy caused by ignoring the misaligned relationship between temperature and DNI. Summary of the Invention

[0005] The purpose of the embodiments of the present application is to provide a data processing method, device, and electronic equipment for a solar thermal power generation system to solve the problem of low DNI prediction accuracy caused by ignoring the misaligned relationship between air temperature and DNI.

[0006] To solve the above technical problems, the first aspect of this specification provides a data processing method for a solar thermal power generation system, comprising:

[0007] Obtain historical meteorological data and historical DNI data of the target CSP system;

[0008] Determining, based on the historical temperature data and the historical DNI data in the historical meteorological data, a daily variation relationship of a hysteresis range between the temperature data and the DNI data over time, wherein the daily variation relationship includes a hysteresis range between the temperature data and the DNI data within a plurality of preset time periods each day;

[0009] Correcting the historical temperature data based on the diurnal variation relationship to obtain target temperature data;

[0010] The historical temperature data in the historical meteorological data is replaced with the target temperature data, so that the processed historical meteorological data and the historical DNI data are used as input data for model training or DNI parameter prediction.

[0011] In some embodiments of the present specification, the hysteresis range between the air temperature data and the DNI data includes a forward expansion amount of the DNI data and a backward compression amount of the air temperature data;

[0012] Accordingly, based on the historical temperature data and the historical DNI data in the historical meteorological data, determining the daily variation relationship of the hysteresis range between the temperature data and the DNI data over time includes:

[0013] constructing a temperature sequence and a DNI sequence based on the historical temperature data and the historical DNI data;

[0014] determining the local rate of change of the temperature series relative to the DNI series over a plurality of preset time periods;

[0015] Based on multiple local change rates, preset adjustment coefficients, and preset expansion amounts, determining the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence within each preset time period;

[0016] Based on the forward expansion of the DNI series and the backward compression of the temperature series within multiple preset time periods, the daily variation relationship of the lag range between the temperature data and the DNI data over time is determined.

[0017] In some embodiments of the present specification, determining the local change rate of the air temperature sequence relative to the DNI sequence within a plurality of preset time periods includes:

[0018] For any preset time period, based on the data characteristics of the air temperature sequence and the DNI sequence within the preset time period, the preset time period is divided into multiple time sub-segments, and the local change rate of the air temperature sequence relative to the DNI sequence within each time sub-segment is determined;

[0019] Accordingly, based on multiple local change rates, preset adjustment coefficients, and preset expansion amounts, the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence within each preset time period are determined, including:

[0020] For any preset time period, based on multiple local change rates, preset adjustment coefficients, and preset expansion amounts, determine the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence in each time sub-segment of the preset time period;

[0021] Accordingly, based on the forward expansion of the DNI series and the backward compression of the temperature series within multiple preset time periods, the daily variation relationship of the lag range between the temperature data and the DNI data over time is determined, including:

[0022] For any preset time period, the dynamic hysteresis range between the temperature parameter and the DNI parameter within the preset time period is determined based on the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence within each time sub-segment;

[0023] The diurnal variation relationship is determined based on dynamic hysteresis ranges of a plurality of preset time periods.

[0024] In some embodiments of the present specification, the daily variation relationship of the hysteresis range between the temperature data and the DNI data over time is expressed by the following formula:

[0025] w low =w base low · (1 +γ·r ( t ));

[0026] w high =w base high · (1 -δ·r ( t ));

[0027] r(t)=|ΔT(t)| / |ΔD(t)|, ΔT(t)=T(t)-T(t-1), ΔD(t)=D(t)-D(t-1);

[0028] Among them, w low Indicates the forward expansion of DNI data, w high Indicates the backward compression of temperature data, w base low Indicates the forward reference extension of DNI data, w base high represents the backward reference compression of the temperature data, γ and δ represent the adjustment coefficients, r(t) represents the local change rate, T(t) and T(t-1) represent the historical temperature data at time t and time t-1, respectively, and D(t) and D(t-1) represent the historical DNI data at time t and time t-1, respectively.

[0029] In some embodiments of the present specification, the plurality of preset time periods include an early stage, a middle stage, and a late stage.

[0030] In some embodiments of the present specification, the forward expansion amount of the DNI data in the early stage is less than the backward compression amount of the temperature data; the forward expansion amount of the DNI data in the late stage is the same as the backward compression amount of the temperature data; the forward expansion amount of the DNI data in the mid-stage and the backward compression amount of the temperature data change with time.

[0031] In some embodiments of the present specification, the forward expansion amount of the DNI data and the backward compression amount of the temperature data in the mid-term stage are respectively expressed by the following formulas:

[0032] w’ low =α· (1 -t' / T' ) +γ·r ( t’ );

[0033] w’ high =β· ( t' / T' ) -δ·r ( t’ );

[0034] r ( t’ )=|Δ T ( t’ )| / |Δ D ( t’ )|,Δ T ( t’ )= T ( t’ )- T ( t’- 1), Δ D ( t’ )= D ( t’ )- D ( t’- 1);

[0035] in, α and β Respectively represent the preset parameters of the mid-term stage, w’ low Indicates the forward expansion of DNI data in the mid-term period, w’ high Indicates the amount of backward compression of the temperature data in the mid-term period, γ and δ represents the adjustment coefficient, r ( t’ ) represents the local rate of change, T ( t’ )and T ( t’- 1) Respectivelyt’ Moment and t’ -1 moment historical temperature data, D ( t’ )and D ( t’- 1) Indicates t’ Moment and t’ -1 moment of historical DNI data, t’ is the current time step or time index, T’ is the total number of time steps or the size of the total time window.

[0036] In some embodiments of the present specification, correcting the historical temperature data based on the diurnal variation relationship to obtain target temperature data includes:

[0037] Determining bandwidth constraints corresponding to temperature parameters and DNI parameters within each preset time period based on the diurnal variation relationship;

[0038] The historical temperature data is corrected based on the bandwidth constraints, historical temperature data, and historical DNI data corresponding to each preset time period to obtain target temperature data.

[0039] In some embodiments of this specification, determining the bandwidth constraints corresponding to the temperature parameter and the DNI parameter within each preset time period based on the diurnal variation relationship includes:

[0040] The bandwidth constraint of each preset time period is determined based on the forward expansion amount of the DNI data and the backward compression amount of the temperature data corresponding to each preset time period, as well as the sequence length corresponding to the DNI data.

[0041] In some embodiments of the present specification, correcting the historical temperature data based on the bandwidth constraints, historical temperature data, and historical DNI data corresponding to each preset time period to obtain target temperature data includes:

[0042] constructing a temperature sequence and a DNI sequence based on the historical temperature data and the historical DNI data;

[0043] For any preset time period, based on the bandwidth constraints of each preset time period and the DNI sequence corresponding to each preset time period, the temperature sequence is processed to obtain multiple temperature subsequences;

[0044] Calculate the matching path scores of multiple temperature subsequences and DNI sequences within each preset time period;

[0045] The temperature subsequence corresponding to the minimum value of the multiple matching path scores in each preset time period is determined as the target temperature sequence of the preset time period, and the target temperature sequence is used as the target temperature data.

[0046] In some embodiments of the present specification, the matching path score is determined by the following formula:

[0047] ;

[0048] ;

[0049] ;

[0050] in, DP ( i , j ) indicates the DNI sequence D ( i ) and the temperature subsequence T ( j ), C ( i , j ) can represent the DNI sequence with the penalty term introduced D ( i ) and the temperature subsequence T ( j ), cost ( D ( i ), T ( j ) indicates the DNI sequence D ( i ) and the temperature subsequence T ( j ), λ represents the penalty coefficient, penalty ( i , j ) represents the penalty function, w low Indicates the forward expansion of DNI data. w high Indicates the amount of backward compression of the temperature data.

[0051] In some embodiments of this specification, the processed historical meteorological data and historical DNI data are used as input data for model training or DNI parameter prediction, including:

[0052] Based on the processed historical meteorological data and historical DNI data, the input data matrix is ​​constructed;

[0053] The input data matrix is ​​used as training data, and a neural network model is trained and tested based on the training data to obtain a DNI parameter prediction model; alternatively, the input data matrix is ​​input into a pre-trained DNI parameter prediction model to output the DNI parameters at the target time point.

[0054] A second aspect of this specification provides a data processing device for a solar thermal power generation system, comprising:

[0055] An acquisition module is used to obtain historical meteorological data and historical DNI data of the target CSP system;

[0056] a determination module configured to determine, based on the historical temperature data and the historical DNI data in the historical meteorological data, a daily variation relationship of a hysteresis range between the temperature data and the DNI data over time, wherein the daily variation relationship includes a hysteresis range between the temperature data and the DNI data within a plurality of preset time periods each day;

[0057] A first processing module is configured to perform correction processing on the historical temperature data based on the diurnal variation relationship to obtain target temperature data;

[0058] The second processing module is used to replace the historical temperature data in the historical meteorological data with the target temperature data, so as to use the processed historical meteorological data and historical DNI data as input data for model training or DNI parameter prediction.

[0059] The third aspect of this specification provides an electronic device, comprising: a memory and a processor, wherein the processor and the memory are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the steps of the method described in the first aspect.

[0060] A fourth aspect of this specification provides a computer storage medium, wherein the computer storage medium stores computer program instructions, and when the computer program instructions are executed, the steps of the method described in the first aspect are implemented.

[0061] A fifth aspect of this specification provides a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of the method described in the first aspect.

[0062] The data processing method, device, and electronic device of a solar thermal power generation system in the embodiments of this specification obtain historical meteorological data and historical DNI data of a target solar thermal power generation system; determine the daily variation relationship of the hysteresis range between the temperature data and the DNI data over time based on the historical temperature data and historical DNI data in the historical meteorological data, wherein the daily variation relationship includes the hysteresis range between the temperature data and the DNI data in multiple preset time periods each day; correct the historical temperature data based on the daily variation relationship to obtain target temperature data; replace the historical temperature data in the historical meteorological data with the target temperature data, and use the processed historical meteorological data and historical DNI data as input data for model training or DNI parameter prediction. The above method provided in the embodiments of this specification determines the daily variation relationship of the hysteresis range between the temperature data and the DNI data over time by analyzing and processing the historical temperature data and the historical DNI data, fully considering the hysteresis relationship between the temperature data and the DNI data in different time periods, that is, the daily variation relationship. Based on the daily variation relationship, dynamic correction processing of the historical temperature data can be achieved, which can optimize the synchronization of the historical temperature data and the historical DNI data and enhance the correlation between the historical temperature data and the DNI data. Furthermore, the input data generated based on the dynamically corrected temperature data can improve the prediction accuracy of the model or the accuracy of DNI parameter prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the implementation methods of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the implementation methods or the description of the prior art. Obviously, the drawings described below are only some implementation methods recorded in this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0064] Figure 1 FIG2 is a schematic diagram of a data processing method for a solar thermal power generation system provided in an embodiment of this specification;

[0065] Figure 2 FIG2 is a schematic diagram of a method for determining a daily variation relationship provided in an embodiment of this specification;

[0066] Figure 3 FIG2 is a schematic diagram of a method for determining target temperature data provided in an embodiment of this specification;

[0067] Figure 4 FIG2 is a schematic diagram of a CNN-Bilstm model provided in an embodiment of this specification;

[0068] Figure 5 FIG2 is a schematic diagram of a DNI prediction method provided in an embodiment of this specification;

[0069] Figure 6 FIG2 is a schematic diagram of a method for calculating a matching path score according to an embodiment of the present disclosure;

[0070] Figure 7 Shown is a schematic diagram of a comparison curve of prediction results provided in the embodiments of this specification;

[0071] Figure 8 FIG2 is a schematic diagram of a data processing device of a solar thermal power generation system provided in an embodiment of this specification;

[0072] Figure 9 Shown is a schematic diagram of an electronic device provided in an embodiment of this specification. DETAILED DESCRIPTION

[0073] In order to enable those skilled in the art to better understand the technical solutions in this application, the following will clearly and completely describe the technical solutions in the embodiments of this application in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0074] It should be noted that the user-related information and data involved in the embodiments of this specification are all information and data authorized by the user or fully authorized by relevant parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data comply with relevant laws, regulations and standards, do not violate public order and good morals, and provide corresponding operation entrances for users or relevant parties to choose to authorize or refuse.

[0075] It should also be noted that in the embodiments of this specification, certain software, components, models and other existing solutions in the industry may be mentioned. They should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of this application, but it does not mean that this application has or will necessarily use the solution.

[0076] Figure 1Shown is a schematic diagram of the data processing method of the solar thermal power generation system provided in the embodiment of this specification. Although this specification provides the method operation steps or device structure as shown in the following embodiments or drawings, the method or device may include more or fewer operation steps or module units after partial merger based on routine or no creative labor. In the steps or structures where there is no necessary causal relationship logically, the execution order of these steps or the module structure of the device is not limited to the execution order or module structure shown in the embodiments or drawings of this specification. When the method or module structure described is applied to actual devices, servers or terminal products, it can be executed sequentially or in parallel according to the method or module structure shown in the embodiments or drawings (for example, in a parallel processor or multi-threaded processing environment, or even in a distributed processing, server cluster implementation environment). As Figure 1 As shown, the method may include:

[0077] S101: Acquire historical meteorological data and historical DNI data of a target CSP system.

[0078] It's understood that historical meteorological data can include data related to temperature, air pressure, wind speed, wind direction, and humidity. Since solar radiation, upon reaching the ground, is first absorbed by the surface and converted into heat energy, the surface temperature rises. The rise or fall in temperature lags behind the solar radiation value, causing the measured temperature data to lag behind the DNI data. Therefore, historical temperature data can be corrected to improve the accuracy of subsequent DNI parameter predictions.

[0079] Direct Normal Irradiance (DNI) represents the amount of direct solar radiation per unit area, perpendicular to the direction of the sun's rays, and is a key parameter for the energy capture efficiency of a CSP system. In practice, DNI can be predicted based on meteorological data and DNI data over a specific time period. This prediction can be used to schedule the concentrator field and regulate the thermal storage system, improving grid stability.

[0080] S102: Based on the historical temperature data and the historical DNI data in the historical meteorological data, determine the daily variation relationship of the hysteresis range between the temperature data and the DNI data over time, wherein the daily variation relationship includes the hysteresis range between the temperature data and the DNI data in multiple preset time periods each day.

[0081] It is understood that in DNI parameter prediction model training or DNI parameter prediction scenarios, it is necessary to consider the temporal changes in meteorological data and DNI data. Therefore, after step S101, a historical meteorological sequence and a historical DNI sequence for the object can be generated based on the acquired historical meteorological data and historical DNI data to reflect the temporal nature of the data. Specifically, the historical meteorological data can include historical temperature data, historical air pressure data, historical wind speed data, historical wind direction data, and historical humidity data. Accordingly, the historical meteorological sequence can include a historical temperature sequence, a historical air pressure sequence, a historical wind speed sequence, a historical wind direction sequence, and a historical humidity sequence.

[0082] Furthermore, the hysteresis range between the temperature data and the DNI data can be the hysteresis range between the temperature sequence and the DNI sequence, and the hysteresis range can characterize the relative movable range of the temperature sequence and the DNI sequence. Specifically, the hysteresis range can include the hysteresis range of the temperature sequence relative to the DNI sequence, that is, the backward compression amount of the temperature sequence, and the hysteresis range of the DNI sequence relative to the temperature sequence, that is, the forward expansion amount of the DNI sequence. Furthermore, due to diurnal variations, the hysteresis range between the temperature sequence and the DNI sequence will fluctuate. Using a fixed bandwidth constraint to correct the temperature sequence may result in alignment errors or invalid matching problems, and the correction computational complexity may increase due to allowing too many invalid path searches. Therefore, by dividing the daily time into multiple time periods, that is, preset time periods, determining the hysteresis range between the temperature sequence and the DNI sequence in each preset time period, and then dynamically correcting the temperature sequence for each preset time period, the accuracy and reliability of the correction can be improved, and the computational complexity can be reduced.

[0083] S103: Correcting the historical temperature data based on the daily variation relationship to obtain target temperature data.

[0084] Specifically, based on the determined daily variation relationship, the bandwidth constraint corresponding to each preset time period can be determined. This bandwidth constraint can be used to determine the length variation range of the temperature sequence. Then, by matching the historical temperature sequence and the historical DNI sequence within this length variation range, the temperature sequence with the highest matching degree with the historical DNI sequence can be obtained as the target temperature data.

[0085] S104: replacing the historical temperature data in the historical meteorological data with the target temperature data, so as to use the processed historical meteorological data and the historical DNI data as input data for model training or DNI parameter prediction.

[0086] It can be understood that the correction processing process of the above-mentioned historical temperature data can be the processing of training data during the model training process, or it can be the processing of model input data in the scenario of applying the trained model to predict DNI parameters. This specification does not limit this.

[0087] Specifically, using processed historical meteorological data and historical DNI data as input data for model training or DNI parameter prediction can include: constructing an input data matrix based on the processed historical meteorological data and historical DNI data; using the input data matrix as training data, and training and testing a neural network model based on the training data to obtain a DNI parameter prediction model; or inputting the input data matrix into a pre-trained DNI parameter prediction model to output the DNI parameters at the target time point.

[0088] In the embodiments of this specification, historical temperature data and historical DNI data are analyzed and processed to determine the daily variation in the lag range between the temperature data and DNI data over time. This fully considers the lag relationship, i.e., the daily variation, between the temperature data and DNI data over different time periods. Based on this daily variation, dynamic correction processing of the historical temperature data can be performed, optimizing the synchronization between the historical temperature data and the historical DNI data and enhancing the correlation between the two data. Furthermore, input data generated based on the dynamically corrected temperature data can improve the prediction accuracy of the model or the accuracy of DNI parameter prediction.

[0089] It is understood that the methods described herein can be applied to electronic devices, which can refer to electronic devices capable of data calculation, processing, and storage. These electronic devices can be terminals such as PCs (personal computers), tablets, smartphones, wearable devices, and intelligent robots; they can also be servers. A server can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server providing cloud computing services.

[0090] In some embodiments of the present specification, the hysteresis range between the air temperature data and the DNI data may include the forward expansion amount of the DNI data and the backward compression amount of the air temperature data. Figure 2 As shown, based on the historical temperature data and the historical DNI data in the historical meteorological data, determining the daily variation relationship of the hysteresis range between the temperature data and the DNI data over time may include:

[0091] S201: Constructing a temperature sequence and a DNI sequence based on the historical temperature data and the historical DNI data;

[0092] S202: Determine the local change rate of the temperature sequence relative to the DNI sequence within a plurality of preset time periods;

[0093] S203: Determining a forward expansion amount of the DNI sequence and a backward compression amount of the temperature sequence within each preset time period based on multiple local change rates, preset adjustment coefficients, and preset expansion amounts;

[0094] S204: Determine the daily variation relationship of the lag range between the temperature data and the DNI data over time based on the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence within multiple preset time periods.

[0095] It can be understood that the constructed temperature sequence can be represented by one-dimensional data after pre-processing the historical temperature data and sorting them in chronological order, and the constructed DNI sequence can be represented by one-dimensional data after pre-processing the historical DNI data and sorting them in chronological order. For example, the DNI sequence can be recorded as D={D1, D2, ..., D i ,…,D N}, the temperature series can be recorded as T={T1,T2,…,T j ,…,T M}.

[0096] It can be understood that the local change rate can be understood as the proportional relationship between the change in temperature data within a certain time period and the change in DNI data. Based on this local change rate, the change in the lag relationship between the temperature sequence and the DNI sequence over time can be determined. Based on this local change rate, the range by which the temperature sequence lags behind the DNI sequence within the time period can be adjusted to ensure the alignment accuracy of the temperature sequence and the DNI sequence.

[0097] It can be understood that the extent to which the temperature sequence lags behind the DNI sequence can also be related to a preset adjustment coefficient. The preset adjustment coefficient and the local change rate can be combined to determine the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence in each preset time period.

[0098] It can be understood that before determining the lag range within each preset time period, a reference lag range applicable to most data can be determined based on clustering and other analysis and processing of a large amount of data, including a forward reference expansion amount and a backward reference compression amount. Then, after determining the above-mentioned local change rate and the preset adjustment coefficient, the forward reference expansion amount and the backward reference compression amount can be adjusted to obtain the range in which the temperature series lags behind the DNI series within the corresponding preset time period.

[0099] It is understood that the lag range between the temperature sequence and the DNI sequence within each preset time period, i.e., the forward extension of the DNI sequence and the backward compression of the temperature sequence, can be constant or dynamically change over time. Furthermore, the forward extension of the DNI sequence can be equal to or different from the backward compression of the temperature sequence, and the difference between them can also dynamically change over time. Specifically, this lag range can be determined based on the temporal trend of the lag relationship between the temperature sequence and the DNI sequence within the corresponding time period. For example, taking the division of the day into early, mid, and late phases, in the early phase, the temperature sequence may lag significantly behind the DNI sequence, necessitating a relaxation of the DNI sequence's lag range to allow the DNI sequence to move forward. In the late phase, the lag tends to stabilize, requiring a tightening of the lag range to ensure alignment accuracy. In the mid phase, the lag range between the temperature sequence and the DNI sequence varies over time. During phases with less fluctuation, the lag range is tightened to reduce invalid path searches, while during phases with greater fluctuation, the lag range is relaxed to avoid missing valid matches. In other words, during the mid phase, the lag range is dynamically adjusted to more accurately capture the lag relationship.

[0100] It is understandable that in other embodiments, other methods of dividing preset time periods may be used. For each preset time period, the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence within the preset time period may be determined based on the changing characteristics of the fluctuation range of the hysteresis relationship within the preset time period.

[0101] In some embodiments of the present specification, determining the local change rate of the air temperature sequence relative to the DNI sequence within multiple preset time periods may include: for any preset time period, based on data characteristics of the air temperature sequence and the DNI sequence within the preset time period, dividing the preset time period into multiple time sub-segments, and determining the local change rate of the air temperature sequence relative to the DNI sequence within each time sub-segment. Furthermore, determining the forward expansion amount of the DNI sequence and the backward compression amount of the air temperature sequence within each preset time period based on the multiple local change rates, the preset adjustment coefficient, and the preset expansion amount may include: for any preset time period, based on the multiple local change rates, the preset adjustment coefficient, and the preset expansion amount, determining the forward expansion amount of the DNI sequence and the backward compression amount of the air temperature sequence within each time sub-segment of the preset time period. Furthermore, determining the daily variation relationship of the lag range between the air temperature data and the DNI data over time based on the forward expansion amount of the DNI sequence and the backward compression amount of the air temperature sequence within multiple preset time periods may include: for any preset time period, determining the dynamic lag range between the air temperature parameter and the DNI parameter within the preset time period based on the forward expansion amount of the DNI sequence and the backward compression amount of the air temperature sequence within each time sub-segment; and determining the daily variation relationship based on the dynamic lag ranges of multiple preset time periods.

[0102] It is understood that each preset time period, or at least part of it, can be divided into multiple time subsegments. The temporal variation characteristics of the hysteresis range between the temperature series and the DNI series in each time subsegment can be similar, such as large hysteresis range fluctuations or stabilization. Based on this, the preset time period can be divided into multiple time subsegments, and the local change rate corresponding to each time subsegment can be calculated. Furthermore, based on the local change rate, preset adjustment coefficient, and preset expansion amount for each time subsegment, the forward expansion amount of the multiple DNI sequences and the backward compression amount of the temperature series for each time subsegment can be obtained. Based on the forward expansion amount of the multiple DNI sequences and the backward compression amount of the temperature series within each preset time period, the time-varying dynamic hysteresis range within that preset time period can be obtained. Furthermore, based on these multiple dynamic hysteresis ranges, the diurnal variation relationship can be determined. By further refining each preset time period, the hysteresis range within that preset time period can be determined to fully reflect the variation characteristics of the hysteresis relationship within that preset time period, providing a basis for subsequent accurate correction of the temperature series.

[0103] In some embodiments of the present specification, the daily variation relationship of the hysteresis range between the temperature data and the DNI data over time can be expressed by the following formula:

[0104] w low =w base low · (1 +γ·r ( t )) Formula (1)

[0105] w high =w base high · (1 -δ·r ( t )) Formula (2)

[0106] r(t)=|ΔT(t)| / |ΔD(t)|, ΔT(t)=T(t)-T(t-1), ΔD(t)=D(t)-D(t-1) Formula (3)

[0107] Among them, w low It can represent the forward expansion of DNI data, w high It can represent the backward compression of temperature data, w base low It can represent the forward reference extension of DNI data, w base highcan represent the backward reference compression of temperature data, γ and δ can represent adjustment coefficients, r(t) can represent the local change rate, T(t) and T(t-1) can represent the historical temperature data at time t and time t-1 respectively, and D(t) and D(t-1) represent the historical DNI data at time t and time t-1 respectively.

[0108] For example, γ and δ can be set to 0.2 and 0.1 respectively. base low and w base high The forward reference expansion amount and backward reference compression amount can be obtained based on the change characteristics of various lag relationships obtained by data analysis and processing of a large amount of temperature data and DNI data. The forward reference expansion amount and backward reference compression amount can then be adjusted based on the local change rate of each preset time period or time sub-segment and the preset adjustment coefficient to obtain the forward expansion amount and backward compression amount corresponding to the preset time period or time sub-segment.

[0109] In some embodiments of the present specification, the plurality of preset time periods may include an early stage, a middle stage, and a late stage.

[0110] Furthermore, the forward expansion amount of the DNI data in the early stage can be less than the backward compression amount of the temperature data; the forward expansion amount of the DNI data in the late stage can be the same as the backward compression amount of the temperature data; and the forward expansion amount of the DNI data and the backward compression amount of the temperature data in the mid-stage can vary over time. For example, the forward expansion amount of the DNI data in the early stage can be 5, and the backward compression amount of the temperature data can be 15; the forward expansion amount of the DNI data in the late stage and the backward compression amount of the temperature data can be 5; and the forward expansion amount of the DNI data and the backward compression amount of the temperature data in the mid-stage can be determined based on the local change rate of that stage, a preset adjustment coefficient, and a preset expansion amount.

[0111] In some embodiments of the present specification, the forward expansion amount of the DNI data and the backward compression amount of the temperature data in the mid-term stage can be respectively expressed by the following formulas:

[0112] w’ low =α· (1 -t' / T' ) +γ·r ( t’ ) Formula (4)

[0113] w’ high =β· ( t' / T' ) -δ·r ( t’ ) Formula (5)

[0114] r ( t’ )=|Δ T ( t’ )| / |Δ D ( t’ )|,Δ T ( t’ )= T ( t’ )- T ( t’- 1), Δ D ( t’ )= D ( t’ )- D ( t’- 1) Formula (6)

[0115] in, α and β Respectively represent the preset parameters of the mid-term stage, w’ low Indicates the forward expansion of DNI data in the mid-term period, w’ high Indicates the amount of backward compression of the temperature data in the mid-term period, γ and δ represents the adjustment coefficient, r ( t’ ) represents the local rate of change, T ( t’ )and T ( t’- 1) Respectively t’ Moment and t’ -1 moment historical temperature data, D ( t’ )and D ( t’- 1) Indicates t’ Moment and t’ -1 moment of historical DNI data, t’ is the current time step or time index, T’ is the total number of time steps or the size of the total time window, t' / T' It is used to adjust the bandwidth constraint according to the progress of time in the mid-term stage. It can be understood that formula (4) and formula (5) can be obtained by respectively transforming the above formula (1) and formula (2).

[0116] In some embodiments of the present specification, correcting the historical temperature data based on the diurnal variation relationship to obtain target temperature data may include: determining bandwidth constraints corresponding to temperature parameters and DNI parameters in each preset time period based on the diurnal variation relationship; and correcting the historical temperature data based on the bandwidth constraints corresponding to each preset time period, the historical temperature data, and the historical DNI data to obtain the target temperature data.

[0117] In some embodiments of the present specification, determining the bandwidth constraints corresponding to the temperature parameter and the DNI parameter in each preset time period based on the diurnal variation relationship may include: determining the bandwidth constraints for each preset time period based on the forward expansion amount of the DNI data and the backward compression amount of the temperature data corresponding to each preset time period, as well as the sequence length corresponding to the DNI data. Exemplarily, the bandwidth constraint may be expressed as j∈[iw low , i+w high ], where i represents the length of the DNI sequence and j represents the length of the temperature sequence.

[0118] refer to Figure 3 As shown, in some embodiments of this specification, correcting the historical temperature data based on the bandwidth constraints, historical temperature data, and historical DNI data corresponding to each preset time period to obtain target temperature data may include:

[0119] S301: Constructing a temperature sequence and a DNI sequence based on the historical temperature data and the historical DNI data;

[0120] S302: For any preset time period, based on the bandwidth constraint of each preset time period and the DNI sequence corresponding to each preset time period, the temperature sequence is processed to obtain multiple temperature subsequences;

[0121] S303: Calculating matching path scores between multiple temperature subsequences and DNI sequences within each preset time period;

[0122] S304: Determine a temperature subsequence corresponding to a minimum value among a plurality of matching path scores within each preset time period as a target temperature sequence for the preset time period, and use the target temperature sequence as the target temperature data.

[0123] It is understood that before calculating the matching path score, the DNI sequence constructed based on the historical DNI data can be divided according to preset time periods to obtain the DNI sequence corresponding to each preset time period for subsequent matching. The temperature subsequence can be directly processed from the temperature sequence constructed based on the historical temperature data based on the DNI sequence within the corresponding preset time period and the bandwidth constraint.

[0124] Specifically, based on the bandwidth constraints of each preset time period and the DNI sequence corresponding to each preset time period, processing the temperature sequence to obtain multiple temperature subsequences can include: determining a value range for each temperature subsequence based on the bandwidth constraints, where the value range can include the length of the temperature subsequence and the amount of extension before and after the DNI sequence; dividing the temperature sequence based on the value range of each temperature subsequence to obtain multiple temperature subsequences, and the temperature sequence corrected within the preset time period can be selected from the multiple temperature subsequences. Furthermore, each temperature subsequence can be matched with the DNI sequence to calculate a matching path score, and based on the calculated multiple matching path scores, a target temperature sequence can be selected from the multiple temperature subsequences as target temperature data, where the target temperature sequence is based on the temperature sequence corrected for time lag.

[0125] In some embodiments of the present specification, the matching path score may be determined by the following formula:

[0126] Formula (7)

[0127] Formula (8)

[0128] Formula (9)

[0129] in, DP ( i , j ) can represent a DNI sequence D ( i ) and the temperature subsequence T ( j ), C ( i , j ) can represent the DNI sequence with the penalty term introduced D ( i ) and the temperature subsequence T ( j ), cost ( D ( i ), T ( j ) can represent a DNI sequence D ( i ) and the temperature subsequence T ( j ), λ It can represent the penalty coefficient, penalty ( i , j ) can represent the penalty function, w lowIt can represent the forward expansion of DNI data. w high It can represent the amount of backward compression of temperature data.

[0130] It can be understood that when calculating the matching path score, the penalty term is introduced into the dynamic programming to force the alignment path to be within the allowed bandwidth range (i.e., bandwidth constraint). A cost function can be constructed based on the penalty term. Under the premise of satisfying the bandwidth constraint, a recursive relationship between the temperature subsequence and the DNI sequence is dynamically programmed based on the cost function, and this recursive relationship is used as the matching path score between the two. Based on the above formulas (7) to (9), the matching path score between each temperature subsequence and the DNI sequence within each preset time period can be calculated, and the temperature subsequence corresponding to the minimum value among the multiple matching path scores calculated can be selected as the target temperature sequence after correction processing within the preset time period.

[0131] In some embodiments of this specification, the corrected target temperature sequence (or target temperature data) can be used for subsequent training of the DNI parameter prediction model. Specifically, a meteorological sequence is constructed based on historical meteorological data, and the temperature sequence therein is replaced with the corrected target temperature sequence. This sequence is then merged with the DNI sequence constructed based on the historical DNI data and normalized to produce an input data matrix. Furthermore, the input data matrix can be divided into a training set and a test set in a certain ratio, for example, 7:3, for use in training the neural network model.

[0132] In some embodiments of this specification, before training the neural network model, a model structure of a neural network model suitable for DNI parameter prediction can be built first. The neural network model can adopt the CNN-Bilstm model. The structure of the model can refer to Figure 4 As shown. Figure 4 As shown, the CNN-Bilstm model can include a CNN layer, a Bilstm layer, a regularization layer (i.e. Figure 4 Dropout in [ ]), fully connected layers, and output layers. CNN layers can include convolutional layers, ReLu activation layers, flat pooling layers, and flattening layers for feature extraction. BILSTM layers can include forward LSTM and backward LSTM for sequence modeling. This model can have 1 convolutional layer, 64 convolution kernels, and a pooling window size of 3.

[0133] Furthermore, the divided training set and test set can be input into the built CNN-Bilstm model for training and testing, and the training parameters can be set, including the number of training epochs, learning rate, batch training size, forgetting rate, etc. For example, the number of training epochs can be 500, the learning rate can be 0.001, the batch training size can be 24, and the forgetting rate can be 0.5.

[0134] Furthermore, after training the DNI parameter prediction model, the mean absolute error (MAE), root mean square error (RMSE), and determination coefficient R 2 Evaluate the performance of the trained model and measure the deviation between the predicted value and the true value from different angles.

[0135] The embodiment of this specification also provides a DNI prediction method based on improved DTW, referring to Figure 5 As shown, the method can include the de-hysteresis processing of the temperature series and the DNI prediction. It can be understood that this method takes the DNI prediction of a certain solar thermal station as an example and divides each day into the early stage, the middle stage and the late stage.

[0136] refer to Figure 5 As shown in Figure 2, the lag elimination process of the temperature series mainly includes the following steps:

[0137] S1. Collect historical DNI data and meteorological data of CSP stations.

[0138] In the embodiment of this specification, DNI data and meteorological data of a certain solar thermal power station in a certain area in a certain month can be collected. The meteorological data may include temperature data, air pressure data, wind speed data, wind direction data, and humidity data.

[0139] S2. Extract the DNI sequence and use the improved DTW to match the temperature data to obtain the matching scores under different temperature sequences.

[0140] Specifically, the extracted DNI sequence can be recorded as D={D1,D2,…,D i ,…,D N}, the temperature series can be recorded as T={T1,T2,…,T j ,…,T M The segmented asymmetric bandwidth constrained DTW algorithm designed for air temperature series is not suitable for matching with a lag relationship between air temperature and DNI compared to the traditional symmetric bandwidth SC-DTW. It has the problems of resource waste and incorrect matching areas. The segmented asymmetric constrained dynamic bandwidth design can improve the alignment efficiency and accuracy. The improved DTW process can be as follows: Figure 6 Specifically, it may include:

[0141] S21, segmented asymmetric bandwidth constraint, divides the time axis into three stages: early stage, middle stage and late stage, and adopts different asymmetric bandwidth constraints for each stage.

[0142] Understandably, the fixed bandwidth constraints used in traditional DTW (such as the Sakoe-Chiba band) have drawbacks. The lag between temperature and DNI can be larger in the morning and smaller at times like noon. Fixed bandwidths can be too restrictive in some phases, leading to alignment errors, while being too loose in other phases, resulting in invalid matches. Furthermore, fixed bandwidths can allow for excessive invalid path searches in certain phases, increasing computational complexity. Therefore, the timeline is divided into three phases: early, middle, and late, with different asymmetric bandwidth constraints applied to each phase.

[0143] In the early stage, the DNI sequence in the matching process is allowed to expand forward and the temperature sequence is compressed backward and fixed. The bandwidth constraints in the early stage are as follows:

[0144] j∈[iw low1 , i+w high1 ],w low1 =5,w high1 =15 formula (10)

[0145] Among them, w low1 and w high1 They represent the forward expansion of the DNI sequence and the backward compression of the temperature sequence respectively; i represents the time index of the DNI sequence, and j represents the time index of the temperature sequence.

[0146] In the mid-term, the bandwidth needs to be dynamically adjusted to more accurately capture lag relationships and avoid alignment errors caused by fixed bandwidth. In the mid-term, the bandwidth will be tightened during periods of low fluctuation to reduce invalid path searches; in periods of high fluctuation, the bandwidth will be loosened to avoid missing valid matches. The bandwidth constraints in the mid-term are as follows:

[0147] j∈[iw low2 , i+w high2 ] Formula (11)

[0148] w low2 =α· (1 -t' / T' ) +γ·r ( t’ ) Formula (12)

[0149] w high2 =β· ( t' / T' ) -δ·r ( t’ ) Formula (13)

[0150] in, α =10, β= 5, is the preset parameter; γ =0.2, δ= 0.1, is the adjustment coefficient; r ( t’ ) is the local rate of change, used to dynamically adjust the bandwidth, t’ is the current time step or time index, T’ is the total number of time steps or the size of the total time window, t' / T' Used to adjust bandwidth constraints over time during the mid-term phase.

[0151] In the late stage, the matching process needs to strictly limit the lag range to ensure the alignment accuracy. The bandwidth constraints in the late stage are as follows:

[0152] j∈[iw low3 , i+w high3 ],w low3 =w high3 =5 formula (14)

[0153] Since the lag between temperature and DNI may change over time, in the early stages of matching, the range of temperature lag behind DNI may be large, and it is necessary to relax the extension range of DNI and allow DNI to move forward, that is, the above formula (10). In the late stages of matching, the lag tends to be stable, and the bandwidth needs to be tightened to ensure the alignment accuracy, that is, the above formula (14). In the middle stage, the local change rate is used to calculate the local change rate. r ( t’ ) Dynamically correct the bandwidth. The specific formula of the local change rate can refer to the previous formula (6); if r ( t’ ) increases, indicating that the temperature fluctuates violently. w low2 Increased, allowing more compression, w high2 The state bandwidth adjustment in the embodiments of this specification can dynamically adjust the allowed matching range according to the local characteristics of the data.

[0154] S22. Introduce a penalty term in dynamic programming to force the alignment path to be within the allowed bandwidth.

[0155] Formula (15)

[0156] Thus, the cost function C(i, j) is obtained:

[0157] Formula (16)

[0158] Among them, λ is the penalty coefficient, which controls the weight of the penalty term.

[0159] S23. Optimize the DTW matching path score.

[0160] Under the premise of satisfying the bandwidth constraint, the recursive relation DP(i,j) (i.e., matching path score) of the improved DTW dynamic programming can be expressed as:

[0161] Formula (17)

[0162] S3. Extract the temperature data corresponding to the DTW matching path score with the lowest score and reconstruct the dataset.

[0163] Specifically, the optimized DTW is used to match the temperature sequence and the DNI sequence, and the minimum DTW matching path score DP is selected. min Temperature series.

[0164] Continue to refer Figure 5 As shown, the DNI prediction process may specifically include:

[0165] S4. Normalize the dataset features and divide them into training set and test set.

[0166] S5. Build a CNN-Bilstm model for prediction.

[0167] Specifically, the data matrix can be divided into training set and test set in a ratio of 7:3; the constructed CNN-Bilstm model is used for training and testing. The network structure of the CNN-Bilstm model is as follows: Figure 4 The main structural parameters of the CNN-Bilstm model used are: number of convolution layers is 1, number of convolution kernels is 64, pooling window size is 3, number of training epochs is 500, learning rate is 0.001, batch size is 24, and forgetting rate is 0.5.

[0168] S6. Output the final prediction results and evaluate the results.

[0169] Specifically, MAE, RMSE and coefficient of determination R2 can be used to compare and evaluate the prediction results of the DNI prediction model provided in the embodiment of this specification with the CNN-Bilstm without temperature series correction. Figure 7 The comparison of indicators is shown in Table 1 below.

[0170] Table 1

[0171]

[0172] The DNI prediction method provided in the embodiments of this specification takes into account the high correlation between temperature and DNI, making it an important parameter input into the DNI prediction model. However, because the historical temperature series used in traditional DNI prediction using meteorological data lags behind the historical DNI series, the correlation between temperature and DNI is weakened. Therefore, based on the coupling relationship between temperature and DNI, the embodiments of this specification preprocess the predicted input temperature data to enhance its synchronization with DNI and improve the correlation between the two, thereby fundamentally reducing the inherent prediction error caused by the lag in the temperature series.

[0173] Moreover, when preprocessing the temperature data, considering that the traditional SC-DTW uses the Sakoe-Chiba deformed window to add constraints to the DTW regular path, the matching path is symmetrical, and the temperature sequence lags, the symmetrical design will not only result in low computational efficiency and incorrect matching, but also cause a fixed bandwidth that cannot follow the collaborative characteristics between matching sequences, thereby increasing the matching error. In the embodiment of this specification, in the early stage of matching, the improved DTW algorithm relaxes the extension range of DNI and allows DNI to move forward. In the late stage of matching, the lag tends to stabilize, and the improved DTW algorithm narrows the bandwidth range and dynamically adjusts the allowed matching range according to the local characteristics of the data, ultimately improving the matching efficiency and matching accuracy of the algorithm. The embodiment of this specification also uses the improved DTW to find the best matching temperature sequence and replaces the original temperature sequence as the input feature of the prediction model, thereby improving the prediction accuracy of the model.

[0174] Based on the data processing method for a solar thermal power generation system described above, one or more embodiments of this specification also provide a data processing device for a solar thermal power generation system. The device may include an apparatus (including a distributed system), software (application), modules, plug-ins, servers, clients, etc. that utilize the methods described in the embodiments of this specification, combined with the necessary implementation hardware. Based on the same innovative concept, the apparatus in one or more embodiments provided in this specification is described in the following embodiments. Because the implementation solutions to the problems solved by the apparatus are similar to the methods, the implementation of the specific apparatus in the embodiments of this specification can refer to the implementation of the aforementioned methods, and any repetitions will not be repeated. As used below, the terms "unit" or "module" may refer to a combination of software and / or hardware that implements a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and contemplated. Figure 8 The figure shows a schematic diagram of a data processing device for a solar thermal power generation system provided in an embodiment of this specification. Figure 8 As shown, the data processing device 800 of the solar thermal power generation system may include:

[0175] An acquisition module 801 is used to acquire historical meteorological data and historical DNI data of a target CSP system;

[0176] a determination module 802 for determining, based on the historical temperature data and the historical DNI data in the historical meteorological data, a daily variation relationship of a hysteresis range between the temperature data and the DNI data over time, wherein the daily variation relationship includes a hysteresis range between the temperature data and the DNI data within a plurality of preset time periods each day;

[0177] A first processing module 803 is configured to perform correction processing on the historical temperature data based on the diurnal variation relationship to obtain target temperature data;

[0178] The second processing module 804 is configured to replace the historical temperature data in the historical meteorological data with the target temperature data, so as to use the processed historical meteorological data and the historical DNI data as input data for model training or DNI parameter prediction.

[0179] In some embodiments of the present specification, the lag range between the air temperature data and the DNI data may include the forward expansion amount of the DNI data and the backward compression amount of the air temperature data; accordingly, the determination module 802 may be specifically used to: construct an air temperature sequence and a DNI sequence based on the historical air temperature data and the historical DNI data; determine the local change rate of the air temperature sequence relative to the DNI sequence in multiple preset time periods; determine the forward expansion amount of the DNI sequence and the backward compression amount of the air temperature sequence in each preset time period based on multiple local change rates, preset adjustment coefficients, and preset expansion amounts; and determine the daily change relationship of the lag range between the air temperature data and the DNI data over time based on the forward expansion amount of the DNI sequence and the backward compression amount of the air temperature sequence in multiple preset time periods.

[0180] In some embodiments of the present specification, when determining the local change rates of the air temperature sequence relative to the DNI sequence within multiple preset time periods, the determination module 802 may be specifically configured to: for any preset time period, based on the data characteristics of the air temperature sequence and the DNI sequence within the preset time period, divide the preset time period into multiple time sub-segments, and determine the local change rate of the air temperature sequence relative to the DNI sequence within each time sub-segment. Accordingly, when determining the forward expansion amount of the DNI sequence and the backward compression amount of the air temperature sequence within each preset time period based on the multiple local change rates, the preset adjustment coefficient, and the preset expansion amount, the determination module 802 may be specifically configured to: for any preset time period, based on the multiple local change rates, the preset adjustment coefficient, and the preset expansion amount, determine the forward expansion amount of the DNI sequence and the backward compression amount of the air temperature sequence within each time sub-segment of the preset time period. Accordingly, when determining the daily variation relationship of the lag range between the temperature data and the DNI data over time based on the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence within multiple preset time periods, the determination module 802 can be specifically used to: for any preset time period, based on the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence within each time sub-segment, determine the dynamic lag range between the temperature parameter and the DNI parameter within the preset time period; and determine the daily variation relationship based on the dynamic lag ranges of multiple preset time periods.

[0181] In some embodiments of the present specification, the daily variation relationship of the hysteresis range between the temperature data and the DNI data over time can be expressed by the following formula:

[0182] w low =w base low · (1 +γ·r ( t ));

[0183] w high =w base high · (1 -δ·r ( t ));

[0184] r(t)=|ΔT(t)| / |ΔD(t)|, ΔT(t)=T(t)-T(t-1), ΔD(t)=D(t)-D(t-1);

[0185] Among them, w low It can represent the forward expansion of DNI data, w high It can represent the backward compression of temperature data, w baselow It can represent the forward reference extension of DNI data, w base high can represent the backward reference compression of temperature data, γ and δ can represent adjustment coefficients, r(t) can represent the local change rate, T(t) and T(t-1) can represent the historical temperature data at time t and time t-1 respectively, and D(t) and D(t-1) represent the historical DNI data at time t and time t-1 respectively.

[0186] In some embodiments of the present specification, the plurality of preset time periods may include an early stage, a middle stage, and a late stage.

[0187] In some embodiments of the present specification, the forward expansion amount of the DNI data in the early stage may be less than the backward compression amount of the temperature data; the forward expansion amount of the DNI data in the late stage may be the same as the backward compression amount of the temperature data; the forward expansion amount of the DNI data in the mid-stage and the backward compression amount of the temperature data may change over time.

[0188] In some embodiments of the present specification, the forward expansion amount of the DNI data and the backward compression amount of the temperature data in the mid-term stage can be respectively expressed by the following formulas:

[0189] w’ low =α· (1 -t' / T' ) +γ·r ( t’ );

[0190] w’ high =β· ( t' / T' ) -δ·r ( t’ );

[0191] r ( t’ )=|Δ T ( t’ )| / |Δ D ( t’ )|,Δ T ( t’ )= T ( t’ )- T ( t’- 1), Δ D ( t’ )= D ( t’ )- D ( t’- 1);

[0192] in,α and β Respectively represent the preset parameters of the mid-term stage, w’ low Indicates the forward expansion of DNI data in the mid-term period, w’ high Indicates the amount of backward compression of the temperature data in the mid-term period, γ and δ represents the adjustment coefficient, r ( t’ ) represents the local rate of change, T ( t’ )and T ( t’- 1) Respectively t’ Moment and t’ -1 moment historical temperature data, D ( t’ )and D ( t’- 1) Indicates t’ Moment and t’ -1 moment of historical DNI data, t’ is the current time step or time index, T’ is the total number of time steps or the size of the total time window.

[0193] In some embodiments of the present specification, the first processing module 803 can be specifically used to: determine the bandwidth constraints corresponding to the temperature parameters and DNI parameters in each preset time period based on the daily variation relationship; and correct the historical temperature data based on the bandwidth constraints corresponding to each preset time period, historical temperature data, and historical DNI data to obtain target temperature data.

[0194] In some embodiments of the present specification, when determining the bandwidth constraints corresponding to the temperature parameters and DNI parameters in each preset time period based on the diurnal variation relationship, the first processing module 803 can specifically be used to: determine the bandwidth constraints for each preset time period based on the forward expansion amount of the DNI data and the backward compression amount of the temperature data corresponding to each preset time period, as well as the sequence length corresponding to the DNI data.

[0195] In some embodiments of the present specification, when the first processing module 803 corrects the historical temperature data based on the bandwidth constraints, historical temperature data, and historical DNI data corresponding to each preset time period to obtain target temperature data, it can be specifically used to: construct a temperature sequence and a DNI sequence based on the historical temperature data and the historical DNI data; for any preset time period, based on the bandwidth constraints of each preset time period and the DNI sequence corresponding to each preset time period, process the temperature sequence to obtain multiple temperature subsequences; calculate the matching path scores of the multiple temperature subsequences and the DNI sequence within each preset time period; determine the temperature subsequence corresponding to the minimum value among the multiple matching path scores within each preset time period as the target temperature sequence for the preset time period, and use the target temperature sequence as the target temperature data.

[0196] In some embodiments of the present specification, the matching path score may be determined by the following formula:

[0197] ;

[0198] ;

[0199] ;

[0200] in, DP ( i , j ) can represent a DNI sequence D ( i ) and the temperature subsequence T ( j ), C ( i , j ) can represent the DNI sequence with the penalty term introduced D ( i ) and the temperature subsequence T ( j ), cost ( D ( i ), T ( j ) can represent a DNI sequence D ( i ) and the temperature subsequence T ( j ), λ It can represent the penalty coefficient, penalty ( i , j ) can represent the penalty function, w lowIt can represent the forward expansion of DNI data. w high It can represent the amount of backward compression of temperature data.

[0201] In some embodiments of the present specification, the second processing module 804 can be specifically used to: construct an input data matrix based on processed historical meteorological data and historical DNI data; use the input data matrix as training data, train and test a neural network model based on the training data, and obtain a DNI parameter prediction model; or input the input data matrix into a pre-trained DNI parameter prediction model to output the DNI parameters at the target time point.

[0202] The description and functions of the above modules can be understood by referring to the data processing method of the solar thermal power generation system, which will not be repeated here.

[0203] The present application also provides an electronic device, such as Figure 9 As shown, the electronic device may include a processor 901 and a memory 902, wherein the processor 901 and the memory 902 may be connected via a bus or other means. Figure 9 The bus connection is taken as an example.

[0204] The processor 901 may be a central processing unit (CPU). The processor 901 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of these chips.

[0205] The memory 902 is a non-transitory computer-readable storage medium that can be used to store non-transitory software programs, non-transitory computer executable programs and modules, such as the program instructions / modules corresponding to the data processing method of the solar thermal power generation system in the embodiment of the present invention (for example, Figure 8 The processor 901 executes various functional applications and data processing of the processor by running the non-transient software programs, instructions, and modules stored in the memory 902, thereby implementing the data processing method of the solar thermal power generation system in the above method embodiment.

[0206] The memory 902 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created by the processor 901, etc. In addition, the memory 902 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 902 may optionally include a memory remotely located relative to the processor 901, and these remote memories may be connected to the processor 901 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0207] The one or more modules are stored in the memory 902 and, when executed by the processor 901, perform the following data processing method for a solar thermal power generation system:

[0208] Obtain historical meteorological data and historical DNI data of a target solar thermal power generation system; determine the daily variation relationship of the lag range between the temperature data and the DNI data over time based on the historical temperature data and the historical DNI data in the historical meteorological data, wherein the daily variation relationship includes the lag range between the temperature data and the DNI data in multiple preset time periods each day; perform correction processing on the historical temperature data based on the daily variation relationship to obtain target temperature data; replace the historical temperature data in the historical meteorological data with the target temperature data, and use the processed historical meteorological data and historical DNI data as input data for model training or prediction of DNI parameters.

[0209] The specific details of the above electronic device can be understood by referring to the corresponding descriptions and effects in the above method embodiments, and will not be repeated here.

[0210] This specification also provides a computer storage medium, which stores computer program instructions. When the computer program instructions are executed, the steps of the data processing method of the above-mentioned solar thermal power generation system are implemented.

[0211] This specification also provides a computer program product, which includes a computer program. When the computer program is executed, the steps of the method for determining the fracturing operation parameters are implemented.

[0212] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When executed, the program can include the processes in the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD). The storage medium can also include a combination of the above-mentioned types of memory.

[0213] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0214] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions.

[0215] For the convenience of description, the above devices are described as being divided into various units according to their functions. Of course, when implementing this application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0216] From the above description of the embodiments, it can be seen that those skilled in the art can clearly understand that the present application can be implemented by means of software plus the necessary general-purpose hardware platform. Based on this understanding, the technical solution of the present application, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute certain parts of the methods of various embodiments of the present application.

[0217] The present application can be used in a wide variety of general-purpose or specialized computer system environments or configurations, such as personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments that include any of the above.

[0218] The present application may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communications network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0219] The above description is merely a preferred embodiment of this specification and is not intended to limit this specification. Those skilled in the art will readily appreciate that various modifications and variations to the embodiments of this specification are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this specification shall be within the scope of protection of this specification.

Claims

1. A data processing method for a solar thermal power generation system, characterized in that: include: Obtain historical meteorological data and historical DNI data of the target CSP system; Determining, based on the historical temperature data and the historical DNI data in the historical meteorological data, a daily variation relationship of a hysteresis range between the temperature data and the DNI data over time, wherein the daily variation relationship includes a hysteresis range between the temperature data and the DNI data within a plurality of preset time periods each day; Correcting the historical temperature data based on the diurnal variation relationship to obtain target temperature data; Replacing the historical temperature data in the historical meteorological data with the target temperature data, so as to use the processed historical meteorological data and the historical DNI data as input data for model training or DNI parameter prediction; The lag range between the temperature data and the DNI data includes the forward expansion of the DNI data and the backward compression of the temperature data; Accordingly, based on the historical temperature data and the historical DNI data in the historical meteorological data, determining the daily variation relationship of the hysteresis range between the temperature data and the DNI data over time includes: constructing a temperature sequence and a DNI sequence based on the historical temperature data and the historical DNI data; For any preset time period, based on the data characteristics of the air temperature sequence and the DNI sequence within the preset time period, the preset time period is divided into multiple time sub-segments, and the local change rate of the air temperature sequence relative to the DNI sequence within each time sub-segment is determined; For any preset time period, based on multiple local change rates, preset adjustment coefficients, and preset expansion amounts, determine the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence in each time sub-segment of the preset time period; For any preset time period, the dynamic hysteresis range between the temperature parameter and the DNI parameter within the preset time period is determined based on the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence within each time sub-segment; determining the diurnal variation relationship based on dynamic hysteresis ranges of a plurality of preset time periods; The daily relationship between the lag range of temperature data and DNI data over time is expressed by the following formula: w low =w base low · (1 +γ·r ( t ))； w high =w base high · (1 -δ·r ( t ))； r ( t )=|D T ( t )| / |D D ( t )|,D T ( t )= T ( t )- T ( t- 1),D D ( t )= D ( t )- D ( t- 1); in, w low Indicates the forward expansion of DNI data. w high Indicates the amount of backward compression of temperature data, w base low Indicates the forward reference extension of DNI data, w base high represents the backward reference compression of the temperature data, γ and δ represents the adjustment coefficient, r ( t ) represents the local rate of change, T ( t )and T ( t- 1) Respectively t Moment and t -1 moment historical temperature data, D ( t )and D ( t- 1) Indicates t Moment and t -1 historical DNI data.

2. The data processing method of the solar thermal power generation system according to claim 1, characterized in that: The plurality of preset time periods include an early stage, a middle stage, and a late stage.

3. The data processing method for a solar thermal power generation system according to claim 2, characterized in that: The forward expansion of the DNI data in the early stage is less than the backward compression of the temperature data; the forward expansion of the DNI data in the late stage is the same as the backward compression of the temperature data; the forward expansion of the DNI data in the middle stage and the backward compression of the temperature data change with time.

4. The data processing method for a solar thermal power generation system according to claim 3, characterized in that: The forward expansion of DNI data and the backward compression of temperature data in the mid-term period are expressed by the following formulas: w’ low =α· (1 -t' / T' ) +γ·r ( t’ )； w’ high =β· ( t' / T' ) -δ·r ( t’ )； r ( t’ )=|D T ( t’ )| / |D D ( t’ )|,D T ( t’ )= T ( t’ )- T ( t’- 1),D D ( t’ )= D ( t’ )- D ( t’- 1); in, α and β Respectively represent the preset parameters of the mid-term stage, w’ low Indicates the forward expansion of DNI data in the mid-term period, w’ high Indicates the amount of backward compression of the temperature data in the mid-term period, γ and δ represents the adjustment coefficient, r ( t’ ) represents the local rate of change, T ( t’ )and T ( t’- 1) Respectively t’ Moment and t’ -1 moment historical temperature data, D ( t’ )and D ( t’- 1) Indicates t’ Moment and t’ -1 moment of historical DNI data, t’ is the current time step or time index, T’ is the total number of time steps or the size of the total time window.

5. The data processing method for a solar thermal power generation system according to claim 1, characterized in that: Correcting the historical temperature data based on the diurnal variation relationship to obtain target temperature data includes: Determining bandwidth constraints corresponding to temperature parameters and DNI parameters within each preset time period based on the diurnal variation relationship; The historical temperature data is corrected based on the bandwidth constraints, historical temperature data, and historical DNI data corresponding to each preset time period to obtain target temperature data.

6. The data processing method for a solar thermal power generation system according to claim 5, characterized in that: Determining bandwidth constraints corresponding to temperature parameters and DNI parameters within each preset time period based on the diurnal variation relationship includes: The bandwidth constraint of each preset time period is determined based on the forward expansion amount of the DNI data and the backward compression amount of the temperature data corresponding to each preset time period, as well as the sequence length corresponding to the DNI data.

7. The data processing method for a solar thermal power generation system according to claim 5, characterized in that: Correcting the historical temperature data based on the bandwidth constraints, historical temperature data, and historical DNI data corresponding to each preset time period to obtain target temperature data includes: constructing a temperature sequence and a DNI sequence based on the historical temperature data and the historical DNI data; For any preset time period, based on the bandwidth constraints of each preset time period and the DNI sequence corresponding to each preset time period, the temperature sequence is processed to obtain multiple temperature subsequences; Calculate the matching path scores of multiple temperature subsequences and DNI sequences within each preset time period; The temperature subsequence corresponding to the minimum value of the multiple matching path scores in each preset time period is determined as the target temperature sequence of the preset time period, and the target temperature sequence is used as the target temperature data.

8. The data processing method for a solar thermal power generation system according to claim 7, characterized in that: The matching path score is determined by the following formula: ； ； ； in, DP ( i , j ) indicates the DNI sequence D ( i ) and the temperature subsequence T ( j ), C ( i , j ) represents the DNI sequence with penalty term introduced D ( i ) and the temperature subsequence T ( j ), cost ( D ( i ), T ( j )) indicates the DNI sequence D ( i ) and the temperature subsequence T ( j ), λ represents the penalty coefficient, penalty ( i , j ) represents the penalty function, w low Indicates the forward expansion of DNI data. w high Indicates the amount of backward compression of the temperature data.

9. The data processing method for a solar thermal power generation system according to claim 1, characterized in that: The processed historical meteorological data and historical DNI data are used as input data for model training or DNI parameter prediction, including: Based on the processed historical meteorological data and historical DNI data, the input data matrix is ​​constructed; The input data matrix is ​​used as training data, and a neural network model is trained and tested based on the training data to obtain a DNI parameter prediction model; alternatively, the input data matrix is ​​input into a pre-trained DNI parameter prediction model to output the DNI parameters at the target time point.

10. A data processing device for a solar thermal power generation system, characterized in that: include: An acquisition module is used to obtain historical meteorological data and historical DNI data of the target CSP system; a determination module configured to determine, based on the historical temperature data and the historical DNI data in the historical meteorological data, a daily variation relationship of a hysteresis range between the temperature data and the DNI data over time, wherein the daily variation relationship includes a hysteresis range between the temperature data and the DNI data within a plurality of preset time periods each day; A first processing module is configured to perform correction processing on the historical temperature data based on the diurnal variation relationship to obtain target temperature data; a second processing module, configured to replace the historical temperature data in the historical meteorological data with the target temperature data, so as to use the processed historical meteorological data and the historical DNI data as input data for model training or DNI parameter prediction; The lag range between the temperature data and the DNI data includes the forward expansion of the DNI data and the backward compression of the temperature data. Accordingly, the determination module is specifically used to: constructing a temperature sequence and a DNI sequence based on the historical temperature data and the historical DNI data; For any preset time period, based on the data characteristics of the air temperature sequence and the DNI sequence within the preset time period, the preset time period is divided into multiple time sub-segments, and the local change rate of the air temperature sequence relative to the DNI sequence within each time sub-segment is determined; For any preset time period, based on multiple local change rates, preset adjustment coefficients, and preset expansion amounts, determine the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence in each time sub-segment of the preset time period; For any preset time period, the dynamic hysteresis range between the temperature parameter and the DNI parameter within the preset time period is determined based on the forward expansion amount of the DNI sequence and the backward compression amount of the temperature sequence within each time sub-segment; determining the diurnal variation relationship based on dynamic hysteresis ranges of a plurality of preset time periods; The daily relationship between the lag range of temperature data and DNI data over time is expressed by the following formula: w low =w base low · (1 +γ·r ( t ))； w high =w base high · (1 -δ·r ( t ))； r ( t )=|D T ( t )| / |D D ( t )|,D T ( t )= T ( t )- T ( t- 1),D D ( t )= D ( t )- D ( t- 1); in, w low Indicates the forward expansion of DNI data. w high Indicates the amount of backward compression of temperature data, w base low Indicates the forward reference extension of DNI data, w base high represents the backward reference compression of the temperature data, γ and δ represents the adjustment coefficient, r ( t ) represents the local rate of change, T ( t )and T ( t- 1) Respectively t Moment and t -1 moment historical temperature data, D ( t )and D ( t- 1) Indicates t Moment and t -1 historical DNI data.

11. An electronic device, characterized in that: include: A memory and a processor, wherein the processor and the memory are communicatively connected to each other, the memory stores computer instructions, and the processor implements the steps of the method according to any one of claims 1 to 9 by executing the computer instructions.

12. A computer storage medium, characterized in that The computer storage medium stores computer program instructions, and when the computer program instructions are executed, the steps of the method according to any one of claims 1 to 9 are implemented.

13. A computer program product, characterized in that The invention comprises a computer program, which implements the steps of the method according to any one of claims 1 to 9 when the computer program is executed by a processor.

Images