Photovoltaic cost prediction method and device, electronic equipment and readable storage medium

Abstract

The invention discloses a photovoltaic cost prediction method and device, electronic equipment and a readable storage medium, and relates to the technical field of photovoltaic cost, and the method comprises the steps: obtaining the planning capacity and planning regional information of a current photovoltaic project, and selecting a current prediction model corresponding to the planning regional information from a plurality of different preset prediction models; inputting the planned capacity into the current prediction model to obtain a first prediction result; extracting a probability coefficient from the current prediction model, performing interval prediction according to the first prediction result and the probability coefficient to obtain an interval prediction result, and outputting the first prediction result and the interval prediction result as a cost prediction result; wherein in the current prediction model, the probability coefficient is used for testing the prediction accuracy of the current prediction model. The method has the effect of improving the accuracy of cost prediction of the photovoltaic project in the coastal region.

Description

Technical Field

[0001] The present application relates to the field of photovoltaic cost technology, and in particular to a photovoltaic cost prediction method, device, electronic device and readable storage medium. Background Art

[0002] Currently, the number of solar photovoltaic power generation projects is growing rapidly. As targets such as "carbon peak" and "carbon neutrality" approach, photovoltaic projects are increasingly contributing to achieving these goals. For investors, the investment and revenue of photovoltaic projects determine whether the projects will continue to move forward. Therefore, it is necessary to estimate the investment costs of photovoltaic projects. However, due to the higher strength requirements of photovoltaic mounting systems in coastal areas, the investment costs are more complex than those of inland projects. Commonly used indicator estimation methods are not suitable for the special requirements of coastal areas, resulting in poor prediction accuracy. Summary of the Invention

[0003] The purpose of this application is to solve at least one of the technical problems existing in the prior art, and to provide a photovoltaic cost prediction method, device, electronic device and readable storage medium, aiming to improve the accuracy of cost prediction of photovoltaic projects in coastal areas.

[0004] In a first aspect, an embodiment of the present application provides a photovoltaic cost prediction method, comprising: Obtaining planned capacity and planned area information of the current photovoltaic project, and selecting a current prediction model corresponding to the planned area information from a plurality of different preset prediction models; Inputting the planned capacity into the current prediction model to obtain a first prediction result; extracting a probability coefficient from the current prediction model, performing interval prediction based on the first prediction result and the probability coefficient to obtain an interval prediction result, and outputting the first prediction result and the interval prediction result as a cost prediction result; Wherein, in the current prediction model, the probability coefficient is used to test the prediction accuracy of the current prediction model.

[0005] According to the technical solution of the embodiment of the present application, at least the following beneficial effects are achieved: since the geographical characteristics such as wind speed and humidity in coastal areas are more complex than those in inland areas, a current prediction model corresponding to the planned area information is selected from multiple different preset prediction models based on the planned area information of the current photovoltaic project, so that the model used for cost prediction is adapted to the geographical characteristics of the current coastal area, thereby improving the prediction accuracy; in addition, in the current prediction model, the prediction accuracy of the current prediction model is also tested by the probability coefficient, thereby ensuring the high accuracy of the prediction result.

[0006] According to some embodiments of the present application, selecting a current prediction model corresponding to the planned area information from a plurality of different preset prediction models includes: Based on the planned area information, geographical feature information corresponding to the planned area information is determined, and then based on the geographical feature information corresponding to the planned area information, a current prediction model corresponding to the geographical feature information of the current photovoltaic project is selected from a plurality of preset prediction models corresponding to different geographical features.

[0007] According to some embodiments of the present application, the preset prediction model is obtained by: Obtain wind speed data for a plurality of different regions, and determine a wind speed range corresponding to a photovoltaic project in a coastal area based on the wind speed data for the plurality of different regions; Obtaining geographical feature information of a target area within the wind speed range for wind speed data, and a historical photovoltaic project data set, wherein the historical photovoltaic project data set includes planned capacity data and cost data; Based on the geographical feature information of the target area and the historical photovoltaic project data set, a plurality of initial prediction models corresponding to different geographical features are obtained; The initial prediction model is subjected to an accuracy test based on a pre-selected probability coefficient to obtain a preset prediction model that passes the accuracy test.

[0008] According to some embodiments of the present application, after obtaining geographical feature information of a target area within the wind speed range and a historical photovoltaic project data set, the photovoltaic cost prediction method further includes: According to different model generation methods for generating different preset prediction models, the historical photovoltaic project data set is subjected to corresponding data preprocessing.

[0009] According to some embodiments of the present application, the pre-selected probability coefficient is obtained by: Preselecting multiple first probability coefficients based on prediction accuracy requirements; According to the cost range restriction requirement, a probability coefficient is selected from a plurality of the first probability coefficients.

[0010] According to some embodiments of the present application, performing interval prediction based on the first prediction result and the probability coefficient to obtain an interval prediction result includes: An interval prediction is performed based on the first prediction result to obtain a predicted cost range, and the probability coefficient is associated with the predicted cost range to obtain an interval prediction result, wherein the probability coefficient is used to characterize the probability that the actual cost falls within the predicted cost range.

[0011] According to some embodiments of the present application, the current prediction model is a linear regression prediction model, and the probability coefficient is used to perform a t-test on the current prediction model.

[0012] In a second aspect, an embodiment of the present application provides an operation control device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the photovoltaic cost prediction method described in the first aspect above.

[0013] In a third aspect, an embodiment of the present application provides an electronic device comprising the operation control device of the second aspect described above.

[0014] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are used to enable a computer to execute the photovoltaic cost prediction method as described in the first aspect above.

[0015] Other features and advantages of the present application will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present application. The purposes and other advantages of the present application can be achieved and obtained through the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the technical solution of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the technical solution of the present application and do not constitute a limitation on the technical solution of the present application.

[0017] The present application is further described below with reference to the accompanying drawings and embodiments; Figure 1 This is a flow chart of a photovoltaic cost prediction method provided by an embodiment of the present application; Figure 2 This is a flow chart of a photovoltaic cost prediction method provided by another embodiment of the present application; Figure 3 This is a flowchart of a method for obtaining a preset prediction model provided by another embodiment of the present application; Figure 4 This is a flowchart of a method for obtaining a preset prediction model provided by another embodiment of the present application; Figure 5 is a flow chart of a method for obtaining a pre-selected probability coefficient provided by another embodiment of the present application; Figure 6 This is a flowchart of a method for obtaining a preset prediction model provided by another embodiment of the present application; Figure 7 is a schematic diagram of a historical photovoltaic project data set provided by another embodiment of the present application; Figure 8 This is a schematic diagram of an operation control device for executing a photovoltaic cost prediction method provided in one embodiment of the present application. DETAILED DESCRIPTION

[0018] This section will describe the specific embodiments of the present application in detail. The preferred embodiments of the present application are shown in the accompanying drawings. The purpose of the accompanying drawings is to supplement the description of the text part of the specification with graphics, so that people can intuitively and vividly understand each technical feature and overall technical solution of the present application, but it cannot be understood as a limitation on the scope of protection of the present application.

[0019] In the description of this application, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, they cannot be understood as limitations on this application.

[0020] In the description of this application, "several" means one or more, "many" means more than two, "greater than," "less than," and "exceed" are understood to exclude the number itself, while "above," "below," and "within" are understood to include the number itself. The terms "first" and "second" are used solely to distinguish technical features and are not to be construed as indicating or implying relative importance, or as implicitly specifying the number or order of the technical features indicated.

[0021] In the description of this application, unless otherwise clearly defined, terms such as setting, installing, and connecting should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in this application based on the specific content of the technical solution.

[0022] The following further describes various embodiments of the photovoltaic cost prediction method of the present application in conjunction with the accompanying drawings.

[0023] like Figure 1 As shown, Figure 1 This is a flowchart of a photovoltaic cost prediction method provided by an embodiment of the present application. The photovoltaic cost prediction method may include but is not limited to step S110, step S120 and step S130.

[0024] Step S110: obtaining the planned capacity and planned area information of the current photovoltaic project, and selecting a current prediction model corresponding to the planned area information from a plurality of different preset prediction models; Step S120: input the planned capacity into the current prediction model to obtain a first prediction result; Step S130: extracting a probability coefficient from the current prediction model, performing interval prediction based on the first prediction result and the probability coefficient to obtain an interval prediction result, and outputting the first prediction result and the interval prediction result as a cost prediction result; Among them, in the current prediction model, the probability coefficient is used to test the prediction accuracy of the current prediction model.

[0025] It is understandable that the specific region of the photovoltaic project will be determined before the cost forecast is made. In other words, some geographical and cultural factors of the specific region of the project will be considered in the design stage before the cost forecast is made. Therefore, it can be considered that the specific region of the project that has been determined is suitable for the implementation of the photovoltaic project. Understandably, before conducting cost forecasts for photovoltaic projects, the specific location of the project has usually been determined. In other words, the decision on the specific location of the project has fully considered various regional factors, including comprehensive factors such as the geographical environment, climatic conditions, land use status, humanistic and social background, and policies and regulations. In addition, from the perspective of project development logic, the determination of the region is the basic premise for cost forecasting, because the cost structure of photovoltaic projects depends on the specific characteristics of the site. For example, the topography directly affects the difficulty and scale of civil engineering projects, solar radiation resources affect power generation efficiency and component selection, and local infrastructure conditions, such as grid access distance and road transportation capacity, will affect electrical engineering and logistics costs. In addition, the specific regional characterization of the project has been determined and the core links of the preliminary feasibility study have been completed, including geological surveys, light data analysis, environmental impact assessments, and local policy compliance reviews. All of these regional attributes can be converted into specific parameter inputs in the subsequent cost forecast model. For example, higher anti-corrosion grade costs may be included for coastal areas.

[0026] Based on this, when the project enters the cost forecasting stage, the rationality of the specific region of the project has been verified in multiple dimensions, including technical feasibility, economic comparison and social acceptance analysis. It can be considered that the specific region of the project that has been determined is suitable for the implementation of the photovoltaic project; therefore, the planned capacity and planned area information of the photovoltaic project can be directly obtained, among which the planned capacity of the photovoltaic project refers to the designed installation capacity for the current photovoltaic project, and the planned area information refers to the planned installation area of ​​the current photovoltaic project. In another embodiment, the planned area information can also include the planned installation areas of different photovoltaic modules / photovoltaic strings, as well as inverters, combiners and other equipment.

[0027] It should be noted that the current photovoltaic project may refer to a photovoltaic project including multiple photovoltaic power generation units in the same region, or it may refer to one power generation unit among multiple photovoltaic power generation units in the same region.

[0028] It is understood that based on the mapping relationship between regional characteristics and model applicability, a current prediction model that matches the planned regional information can be selected from multiple preset prediction models, where the multiple preset prediction models are of different types. Specifically, the prediction model can be a machine learning-based regression model (such as random forest, gradient boosting tree, or neural network), a traditional engineering estimation method (unit capacity cost method, bill of quantities method), or a hybrid model (combining data-driven and engineering empirical formulas). The model selection mechanism is based on the structured processing and matching logic of regional information. For example, a regional feature database can be constructed to encode key variables in the planned regional information, such as geographic parameters (altitude, slope, solar radiation, etc.), climate conditions (average annual temperature, wind speed, precipitation, etc.), infrastructure (grid access distance, road conditions, etc.), and policy environment (subsidy intensity, land cost, etc.). These variables are then matched against the training backgrounds of different prediction models to automatically select the optimal model. Prediction model matching can be implemented using a rule engine or machine learning classifier. In another embodiment, the correspondence between the models can also be achieved through dynamic weight adjustment, that is, the output results of multiple models are weighted and fused according to the correlation between regional characteristics and the historical accuracy of each model. For example, in distributed photovoltaic projects, a small-scale system-specific installation cost model may be used for urban areas with dispersed rooftop resources, while a large-scale civil engineering and transmission cost algorithm is used for centralized power stations in uninhabited areas.

[0029] In this embodiment, a linear regression analysis prediction model can be used to predict the construction cost, wherein multiple different preset prediction models are all linear regression analysis prediction models, and multiple different linear regression analysis prediction models are established based on different historical data. Therefore, multiple different linear regression analysis prediction models correspond to different planning area information.

[0030] After selecting the current prediction model, the planned capacity can be input into the current prediction model to obtain the first prediction result; it can be understood that in the current prediction model, the nonlinear relationship between key variables such as regional characteristics and capacity scale represented by the planned regional information has been quantitatively mapped in advance. Among them, the planned capacity, as the core variable driving the cost, is directly related to the system configuration scale, the number of components, the land occupied area, etc., and thus affects the main cost components such as equipment procurement, civil engineering, and electrical access. Therefore, the planned capacity is input into the current prediction model to obtain the first prediction result. In this embodiment, the first prediction result is a point prediction result, that is, a single value estimate of the total project cost under a certain confidence level. The first prediction result can be a predicted cost value or multiple predicted cost values.

[0031] In another embodiment, the output of the first prediction result may not be a static process, but may be dynamically optimized along with the real-time parameter adjustment mechanism of the model. For example, when the input planning capacity exceeds the model training range, the extrapolation algorithm may be triggered or switched to the backup sub-model to ensure the rationality of the prediction.

[0032] It is understandable that, considering that in actual work, the actual construction cost is almost impossible to be completely consistent with the predicted value, a rough range of values ​​will generally be provided to the construction parties, and the construction parties can roughly estimate the project risks based on the range of values.

[0033] Therefore, in this embodiment, a probability coefficient can be extracted from the current prediction model, and an interval prediction can be performed based on the first prediction result and the probability coefficient to obtain an interval prediction result. In this embodiment, the probability coefficient refers to a coefficient extracted from the current prediction model that can reflect the uncertainty of the construction cost. The probability coefficient can include the distribution of historical prediction errors, variable sensitivity weights, or probabilistic indicators output by the model itself (such as the variance of random forest predictions and uncertainty estimates of neural networks). In one embodiment, the interval prediction method can be to use the probability coefficient to probabilistically correct the first prediction result. For example, when the model's built-in error follows a normal distribution, the probability coefficient may be reflected as a standard deviation or the width of a confidence interval. In this case, a 95% confidence interval can be generated by superimposing the point prediction value with plus or minus 1.96 times the standard deviation.

[0034] In one embodiment, the interval forecast is also adjusted based on external risk factors. For example, when the probability coefficient includes the risk of climate fluctuations specific to the region (such as the frequency of typhoons in coastal areas), the cost offset caused by extreme events can be further superimposed on the basic error distribution. In addition, the probability coefficient may also include an engineering experience coefficient. For example, the fluctuation range of civil engineering costs for projects in coastal areas can be manually adjusted, so that the final interval forecast includes both data-driven statistical uncertainty and the subjective judgment of experts on regional specificity.

[0035] It is understandable that in the current prediction model for photovoltaic projects, the probability coefficient can effectively test the accuracy of the current prediction model by statistically analyzing the deviation between the model's historical prediction results and the actual cost. The specific implementation process is as follows: First, during the training phase, the current prediction model will record the key features of each prediction case (project capacity, regional type, component technology route, etc.) and the deviation between the corresponding predicted value and the actual value. After probabilistic processing, the deviation data can be formed into a probability coefficient, which is usually the error distribution law under different scenarios. For example, the cost prediction error of a project on flat land in a coastal area may follow a normal distribution with a mean of 3% and a standard deviation of 2%, while the error of a project on a rooftop in a coastal area may show a right-skewed characteristic; Based on this, we can match the corresponding probability coefficients according to attributes such as regional characteristics, and judge the reliability of the current prediction by comparing the error distribution of the current prediction results with similar historical cases. For example, if the predicted cost of a coastal photovoltaic project is 800 million yuan, and the probability coefficient of similar historical projects shows that 95% of the cases have an error within ±5%, then it can be determined that the prediction has a high confidence level; if the current prediction value deviates from the historical error band (such as falling outside the ±10% range), an accuracy warning will be triggered, prompting the need for manual review of input parameters or switching of the prediction model. In addition, the probability coefficient can also dynamically optimize the model weight. For example, when the recent actual cost of a coastal salt spray area continues to be higher than the predicted value, the probability coefficient of the category will be automatically adjusted to expand its error range, so that the accuracy fluctuations of the model in this scenario will be more realistically reflected in subsequent predictions.

[0036] like Figure 2 As shown, Figure 2 This is a flowchart of a photovoltaic cost prediction method provided by another embodiment of the present application; regarding the above-mentioned step S110, it may include but is not limited to step S210.

[0037] Step S210: Determine geographic feature information corresponding to the planned area information based on the planned area information, and then select a current prediction model corresponding to the planned area information of the current photovoltaic project from a plurality of preset prediction models corresponding to different geographic features based on the geographic feature information corresponding to the planned area information.

[0038] It can be understood that in this embodiment, when selecting a preset prediction model based on the planned area information, the selection can be mainly based on the geographical feature information in the planned area information, that is, the current prediction model corresponding to the geographical feature information of the current photovoltaic project is selected from multiple preset prediction models corresponding to different geographical features.

[0039] For example, we can first perform a structured analysis of the geographical features represented by the planning area information, extract parameters such as terrain type, altitude, slope and aspect, surface cover, and solar radiation intensity distribution, and standardize the parameters to form a geographical feature coding vector; then, we can compare the geographical feature coding vector with the geographical feature labels adapted to each medical model in the preset model library for similarity. For example, we can use the cosine similarity algorithm or rule-based feature matching to select the prediction model with the highest degree of geographical feature consistency as the prediction model for the current project.

[0040] like Figure 3 As shown, Figure 3 This is a flowchart of a photovoltaic cost prediction method provided by another embodiment of the present application; the method for obtaining the preset prediction model may include but is not limited to step S010, step S020, step S030 and step S040.

[0041] Step S010: Obtain wind speed data for a plurality of different regions, and determine the wind speed range corresponding to the photovoltaic project in the coastal area based on the wind speed data for the plurality of different regions; Step S020: Obtaining geographical feature information of a target area within a wind speed range and a historical photovoltaic project data set, the historical photovoltaic project data set including planned capacity data and cost data; Step S030: Based on the geographical feature information of the target area and the historical photovoltaic project data set, a plurality of initial prediction models corresponding to different geographical features are obtained; Step S040: Perform an accuracy test on the initial prediction model based on the pre-selected probability coefficient to obtain a preset prediction model that passes the accuracy test.

[0042] In this embodiment, the multiple preset prediction models are models of the same type, for example, a regression model or a linear regression model based on machine learning.

[0043] It is understandable that wind speed is a key geographical factor that distinguishes inland from coastal areas, and it is also a key factor affecting the construction cost of photovoltaic projects. This is because higher wind speeds increase the wind pressure load on photovoltaic modules, requiring the mounting system to use higher-strength materials and a more stable foundation design, which directly leads to increased mounting and installation costs. In addition, strong wind areas may require additional wind protection measures, such as adding counterweights or adopting special wind-resistant mounting structures, which will increase the initial investment cost of the project. At the same time, frequent strong winds may also affect the construction progress, increase the standby time of labor and machinery, and indirectly increase the total project cost. Therefore, it is necessary to first obtain wind speed data for various different regions, including the annual average wind speed, maximum wind speed and its frequency in coastal areas, and the normal wind speed distribution in inland plains.

[0044] By analyzing wind speed data from a variety of different regions, the boundary range of the photovoltaic project corresponding to the coastal area is determined. In this embodiment, the boundary range represents the wind speed of the area to which the historical data used to generate the model belongs, and should be within the boundary range of the photovoltaic project corresponding to the coastal area. Because the geographical characteristics of the wind speed data within the boundary range of the photovoltaic project corresponding to the coastal area have data value for analyzing the difference in engineering costs between inland areas and coastal areas.

[0045] For example, by using historical wind speed data for coastal and inland areas, including key indicators such as annual average wind speed, extreme wind speed and its frequency of occurrence, and seasonal variation characteristics, the cumulative distribution curves of wind speed in coastal and inland areas are drawn, and the interval where the two curves are most obviously separated is found, thereby obtaining the first wind speed range; or using machine learning algorithms such as support vector machines to classify and train wind speed data, the first wind speed range that can best distinguish coastal and inland characteristics is found; then, considering the actual significance of the project, that is, the wind speed that will indeed have a substantial impact on the cost of photovoltaic projects, the boundary range of photovoltaic projects in coastal areas is obtained. For example, the frequency of wind speeds in the range of 6-12 meters per second in coastal areas is higher than that in inland areas, and this wind speed range is the range where the design of photovoltaic brackets needs to pay special attention to wind resistance. Therefore, the boundary range of photovoltaic projects in coastal areas can not only effectively distinguish regional types, but also have a wind speed range that has a practical impact on project costs, and has data value for analyzing the cost differences between inland and coastal areas.

[0046] Therefore, based on the database currently stored for model training, the wind speed data can be queried for areas within the wind speed range of photovoltaic projects in corresponding coastal areas to determine the target area, and then the geographical feature information of the target area and the historical photovoltaic project data group can be obtained, where the historical photovoltaic project data group includes planned capacity data and cost data.

[0047] It is foreseeable that the target region includes multiple different regions, and therefore, the geographical features corresponding to the multiple different regions are different.

[0048] Subsequently, model training can be performed based on the geographical feature information of multiple different regions in the target area and the historical photovoltaic project data set. For example, multiple models can be trained based on the geographical feature information and historical photovoltaic project data set corresponding to each region to obtain multiple initial prediction models corresponding to different geographical features. It can be understood that when training the cost prediction model, historical planned capacity data is used as input data and cost data is used as output data. Geographic feature information can have multiple functions. For example, as a basic classification variable, data samples can be directly divided by regional type labels, so that the model can identify cost patterns in different geographical environments; as a continuous explanatory variable, quantitative indicators such as altitude, slope, and wind speed are included in the regression analysis to obtain the nonlinear relationship between these factors and cost; as an interactive feature, it is combined with engineering parameters such as planned capacity (capacity × wind speed coefficient) to reflect the regulatory effect of geographical conditions on scale effects.

[0049] At this time, the initial prediction model has not been verified by data. Therefore, the accuracy of the initial prediction model can be tested based on the pre-selected probability coefficient. For example, a probability coefficient library containing prediction error distribution characteristics under different regional characteristics, climatic conditions and technical parameter combinations is obtained in advance through historical data analysis. After the initial prediction model generates a cost estimate, the probability coefficient corresponding to the current project characteristics is automatically matched, and the model prediction value is compared with the historical actual cost data of this type of project for probability. The model accuracy is evaluated by calculating the probability that the prediction value falls within the historical error confidence interval. For example, if the initial prediction result of a coastal project is 85 million yuan, and the probability coefficient of this type of project shows that the historical 95% confidence interval is 80 million to 90 million, it can be determined that the prediction model has passed the basic accuracy test; if the prediction value exceeds the reasonable fluctuation range, the model correction process can be triggered until it finally passes the basic accuracy test; in this way, the final preset prediction model that passes the accuracy test is obtained.

[0050] like Figure 4 As shown, Figure 4 This is a flowchart of a photovoltaic cost prediction method provided by another embodiment of the present application; the method for obtaining the above-mentioned prediction model may also include but is not limited to step S050.

[0051] Step S050: performing data preprocessing on the historical photovoltaic project data set in a corresponding manner according to different model generation methods for generating different preset prediction models.

[0052] In this embodiment, after obtaining the geographical feature information of the target area within the wind speed range and the historical photovoltaic project data set, data preprocessing can be performed on these historical photovoltaic project data sets. If different preset prediction models are models of different types and structures, the current methods used to generate different preset prediction models are different model generation methods, and the historical photovoltaic project data sets can be preprocessed in corresponding ways. For example, the highest total cost value, the lowest total cost value, or the highest unit cost value, the lowest unit cost value, etc. can be removed. If different preset prediction models are models of the same type and structure, the current methods used to generate different preset prediction models can be the same model generation method, and the historical photovoltaic project data sets can be preprocessed in corresponding ways using the same data preprocessing method. In addition, if the preset prediction model is also a linear regression prediction model, preprocessing can be omitted because a regression test will be performed after the curve is subsequently fitted. If there is no causal relationship between the historical data or there are errors in the historical data, the regression test will not pass.

[0053] like Figure 5 As shown, Figure 5This is a flowchart of a photovoltaic cost prediction method provided by another embodiment of the present application; the method for obtaining the above-mentioned pre-selected probability coefficient may include but is not limited to step S060 and step S070.

[0054] Step S060: preselect multiple first probability coefficients according to the prediction accuracy requirement; Step S070: Select a probability coefficient from a plurality of first probability coefficients according to the cost range restriction requirement.

[0055] In this embodiment, based on the prediction accuracy requirements, multiple first probability coefficients that meet different levels of prediction accuracy requirements can be pre-selected by performing hierarchical statistics and feature extraction on historical prediction errors. For example, an analysis library containing a large amount of historical project error data can be constructed, wherein the historical project error data is classified and stored according to dimensions such as project type, regional characteristics, and installed capacity. Then, probability distribution fitting is performed on the prediction errors under each classification, and key statistics at different confidence levels are calculated. For example, for the coastal project subcategory of ground-based centralized photovoltaic power stations, it may be calculated that its error follows the interval distribution of [-4.2%, +5.8%] at an 80% confidence level, and expands to [-7.5%, +9.3%] at a 95% confidence level. The boundary values ​​of these two intervals form the probability coefficients corresponding to different accuracy requirements.

[0056] Subsequently, according to the cost range restriction requirement, a probability coefficient is selected from multiple first probability coefficients. For example, a probability coefficient can be selected from multiple first probability coefficients based on the results of experimental prediction. For example, the cost range restriction requirement requires that the total cost must be controlled between 800 million and 1 billion yuan. At this time, all first probability coefficients that can make the prediction result fall within this range can be screened out, and then the confidence levels corresponding to these first probability coefficients and their engineering rationality can be analyzed. For example, although a certain first probability coefficient can constrain the prediction value within the target range, its corresponding confidence level is only 70%, which means that there is a 30% risk of exceeding the limit. At this time, the coefficient that can achieve the highest confidence level and still meet the cost limit can be selected as the probability coefficient.

[0057] like Figure 6 As shown, Figure 6 This is a flowchart of a photovoltaic cost prediction method provided by another embodiment of the present application; regarding the above-mentioned step S130, it may include but is not limited to step S230.

[0058] Step S230: perform interval prediction based on the first prediction result to obtain a predicted cost range, associate the probability coefficient with the predicted cost range to obtain an interval prediction result, wherein the probability coefficient is used to represent the probability that the actual cost falls within the predicted cost range.

[0059] In this embodiment, interval prediction is performed based on the first prediction result, that is, upward prediction and downward prediction are performed based on the first prediction result. For example, the first prediction result is 20.3 million yuan. The interval prediction is performed based on the first prediction result, and the predicted cost range is (1637, 2423). The first prediction result in the predicted cost range can be the midpoint of the predicted cost range, or it can not be the midpoint.

[0060] In addition, in the current prediction model, the probability coefficient is used to test the prediction accuracy of the current prediction model. The probability coefficient can be an error rate requirement. For example, when the probability coefficient is 0.05, it indicates that the prediction error rate of the current prediction model needs to be less than 0.05. Based on this, the probability coefficient can also be used as an error rate description for the predicted cost range, that is, the probability coefficient is associated with the predicted cost range to obtain an interval prediction result, wherein the probability coefficient is used to represent the probability that the actual cost falls within the predicted cost range. For example, when the probability coefficient is 0.05, it indicates that the probability that the actual cost does not fall within the predicted cost range is 0.05, and the probability that the actual cost falls within the predicted cost range is 1-0.05=0.95.

[0061] In a photovoltaic cost prediction method provided in another embodiment of the present application, the current prediction model is a linear regression prediction model, and the probability coefficient is used to perform a t-test on the current prediction model.

[0062] In this embodiment, the current prediction model is a linear regression prediction model, and the probability coefficient is used to perform a t-test on the current prediction model.

[0063] The following describes the embodiments of the present application in detail by taking the current prediction model as a linear regression prediction model and the probability coefficient as an example for performing a t-test on the current prediction model.

[0064] Exemplarily, the current prediction model is a linear regression prediction model. Regarding the method of obtaining the preset prediction model, first, wind speed data from multiple different regions can be obtained. Based on the wind speed data from multiple different regions, the wind speed range corresponding to the photovoltaic project in the coastal area can be determined. Alternatively, the wind speed range corresponding to the photovoltaic project in the coastal area can be directly determined based on relevant design standards and specifications. Subsequently, geographical feature information of the target area with wind speed data within the wind speed range and a historical photovoltaic project data group are obtained. Among them, the historical photovoltaic project data group corresponding to the target area with wind speed data within the wind speed range is as follows: Figure 7 As shown, Figure 7The data represents historical photovoltaic project data groups in different regions with the same geographical feature information. That is, there are multiple target regions with wind speed data initially acquired within the wind speed range. Each target region corresponds to geographical feature information and a historical photovoltaic project data group. Therefore, classification and grouping can be performed based on the geographical feature information first, and the target regions with the same geographical feature information can be grouped together. Then, model training and generation can be performed based on the historical photovoltaic project data groups corresponding to each group of target regions. In this embodiment, 10 groups of historical photovoltaic project data groups are used as an example, but more historical photovoltaic project data groups can actually be obtained. After obtaining the historical photovoltaic project data set, model training can be performed based on the historical photovoltaic project data set. The basic formula of the linear regression prediction model is y=a+bx. After training with the above historical photovoltaic project data set, a=-66.153 and b=419.205 are obtained, and the initial prediction model can be obtained. This can then be used to train the historical photovoltaic project data set in the target regional group corresponding to each different geographical feature, and multiple initial prediction models corresponding to different geographical features can be obtained. Subsequently, the accuracy of the initial prediction model is tested based on the pre-selected probability coefficient, that is, a t-test is performed, wherein the model t value can be first obtained based on the initial prediction model, and the model t value is calculated to be 13.936 by y=a+bx, a=-66.153, b=419.205. The pre-selected probability coefficient may include a significance level of 0.05 and n-2 degrees of freedom, wherein n is the number of historical photovoltaic project data groups. Then, by querying the t-test boundary value table, the t boundary value corresponding to the horizontal significance level of 0.05 and the vertical freedom degree of 8 is found to be 2.306. When the model t value is greater than the t boundary value, it can be considered to have passed the accuracy test. In this embodiment, the model t value is 13.936, which is greater than the t boundary value of 2.306. Therefore, it can be considered to have passed the accuracy test. In this way, the establishment of the preset prediction model is completed.

[0065] In actual application, the planned capacity and planned area information of the current photovoltaic project are obtained. First, the current prediction model is selected based on the planned area information, and then the planned capacity is input into the current prediction model. In this embodiment, the planned capacity of the current photovoltaic project is 5MWp, and the first prediction result is 20.3 million yuan. Due to changes in the actual situation and the influence of various environmental factors, the actual value will always be offset, so interval prediction is also required. In this embodiment, the probability coefficient is extracted from the current prediction model, and the probability coefficient has a significance level of 0.05. The predicted cost range according to the first prediction result is (16.37 million yuan, 24.23 million yuan), and the probability that the actual value falls within the predicted cost range is 1-0.05=0.95. That is, when the scale of the proposed photovoltaic project in the coastal area is 5MWp, the predicted investment cost will be between 16.37 million yuan and 24.23 million yuan, and the probability that the actual investment falls within this range is 95%.

[0066] Based on the photovoltaic cost prediction methods of the above-mentioned embodiments, various embodiments of the operation control device, electronic device, and computer-readable storage medium of the present application are respectively proposed below.

[0067] like Figure 8 As shown, Figure 8 Schematic diagram of an operation control device for executing a photovoltaic cost prediction method provided by an embodiment of the present application. The operation control device 800 implemented in the present application includes: a processor 820, a memory 810, and a computer program stored in the memory 810 and executable on the processor 820, wherein: Figure 8 In the figure, a processor 820 and a memory 810 are taken as an example.

[0068] The processor 820 and the memory 810 may be connected via a bus or other means. Figure 8 The bus connection is taken as an example.

[0069] The memory 810 is a non-transitory computer-readable storage medium that can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory 810 may include a high-speed random access memory and may also include a non-transitory memory, such as at least one disk storage device, a flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 810 may optionally include a memory 810 remotely located relative to the processor 820, and these remote memories 810 may be connected to the operation control device 800 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.

[0070] Those skilled in the art will understand that Figure 8The device structure shown in the figure does not constitute a limitation on the operation control device 800, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0071] exist Figure 8 In the illustrated operation control device 800, processor 820 can be used to invoke a control program stored in memory 810 to implement the aforementioned photovoltaic cost prediction method. Specifically, the non-transient software program and instructions required to implement the photovoltaic cost prediction method of the aforementioned embodiment are stored in memory 810 and, when executed by processor 820, perform the photovoltaic cost prediction method of the aforementioned embodiment.

[0072] It is worth noting that since the operation control device 800 of the embodiment of the present application can execute the photovoltaic cost prediction method of any of the above-mentioned embodiments, the specific implementation methods and technical effects of the operation control device 800 of the embodiment of the present application can refer to the specific implementation methods and technical effects of the photovoltaic cost prediction method of any of the above-mentioned embodiments.

[0073] In addition, an embodiment of the present application further provides an electronic device, which includes the operation control device of the above embodiment.

[0074] It is worth noting that since the electronic device of the embodiment of the present application includes the operation control device of the above embodiment, and the operation control device of the above embodiment can execute the photovoltaic cost prediction method of any of the above embodiments, the specific implementation methods and technical effects of the electronic device of the embodiment of the present application can refer to the specific implementation methods and technical effects of the photovoltaic cost prediction method of any of the above embodiments.

[0075] In addition, an embodiment of the present application further provides a computer-readable storage medium, which stores computer-executable instructions, and the computer-executable instructions are used to execute the above-described photovoltaic cost prediction method. Figures 1 to 7 The method steps in .

[0076] It is worth noting that since the computer-readable storage medium of the embodiment of the present application can execute the photovoltaic cost prediction method of any of the above embodiments, the specific implementation methods and technical effects of the computer-readable storage medium of the embodiment of the present application can refer to the specific implementation methods and technical effects of the photovoltaic cost prediction method of any of the above embodiments.

[0077] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and appropriate combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on computer-readable media, which can include computer storage media or non-transitory media and communication media or transient media. As is well known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVDs) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, as is well known to those skilled in the art, communication media typically embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0078] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0079] It should also be understood that the various implementation methods provided in the embodiments of the present application can be combined arbitrarily to achieve different technical effects.

[0080] The embodiments of the present application are described in detail above in conjunction with the accompanying drawings, but the present application is not limited to the above embodiments. Various changes can be made within the scope of knowledge possessed by ordinary technicians in the technical field without departing from the purpose of the present application.

Claims

1. A photovoltaic cost prediction method, characterized in that: include: Obtaining planned capacity and planned area information of the current photovoltaic project, and selecting a current prediction model corresponding to the planned area information from a plurality of different preset prediction models; Inputting the planned capacity into the current prediction model to obtain a first prediction result; extracting a probability coefficient from the current prediction model, performing interval prediction based on the first prediction result and the probability coefficient to obtain an interval prediction result, and outputting the first prediction result and the interval prediction result as a cost prediction result; Wherein, in the current prediction model, the probability coefficient is used to test the prediction accuracy of the current prediction model.

2. The photovoltaic cost prediction method according to claim 1, characterized in that: The selecting of a current prediction model corresponding to the planned area information from a plurality of different preset prediction models includes: Based on the planned area information, geographical feature information corresponding to the planned area information is determined, and then based on the geographical feature information corresponding to the planned area information, a current prediction model corresponding to the geographical feature information of the current photovoltaic project is selected from a plurality of preset prediction models corresponding to different geographical features.

3. The photovoltaic cost prediction method according to claim 1, characterized in that: The preset prediction model is obtained in the following way: Obtain wind speed data for a plurality of different regions, and determine a wind speed range corresponding to a photovoltaic project in a coastal area based on the wind speed data for the plurality of different regions; Obtaining geographical feature information of a target area within the wind speed range for wind speed data, and a historical photovoltaic project data set, wherein the historical photovoltaic project data set includes planned capacity data and cost data; Based on the geographical feature information of the target area and the historical photovoltaic project data set, a plurality of initial prediction models corresponding to different geographical features are obtained; The initial prediction model is subjected to an accuracy test based on a pre-selected probability coefficient to obtain a preset prediction model that passes the accuracy test.

4. The photovoltaic cost prediction method according to claim 3, characterized in that: After obtaining geographical feature information of a target area within the wind speed range and a historical photovoltaic project data set, the photovoltaic cost prediction method further includes: According to different model generation methods for generating different preset prediction models, the historical photovoltaic project data set is subjected to corresponding data preprocessing.

5. The photovoltaic cost prediction method according to claim 3, characterized in that: The pre-selected probability coefficient is obtained in the following way: Preselecting multiple first probability coefficients based on prediction accuracy requirements; According to the cost range restriction requirement, a probability coefficient is selected from a plurality of the first probability coefficients.

6. The photovoltaic cost prediction method according to claim 1, characterized in that: The performing interval prediction according to the first prediction result and the probability coefficient to obtain an interval prediction result includes: An interval prediction is performed based on the first prediction result to obtain a predicted cost range, and the probability coefficient is associated with the predicted cost range to obtain an interval prediction result, wherein the probability coefficient is used to characterize the probability that the actual cost falls within the predicted cost range.

7. The photovoltaic cost prediction method according to claim 1, characterized in that: The current prediction model is a linear regression prediction model, and the probability coefficient is used to perform a t-test on the current prediction model.

8. An operation control device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the photovoltaic cost prediction method according to any one of claims 1 to 7.

9. An electronic device, characterized in that: Including the operation control device according to claim 8.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the photovoltaic cost prediction method according to any one of claims 1 to 7.

Images