Photovoltaic power generation energy efficiency analysis method and device suitable for cigarette factory and storage medium

Abstract

The invention discloses a photovoltaic power generation energy efficiency analysis method and device suitable for a cigarette factory and a storage medium, and belongs to the technical field of photovoltaic power generation. The method comprises the following steps: acquiring historical generating capacity data and real-time monitoring data of a photovoltaic power generation system; inputting the historical generating capacity data and the real-time monitoring data into the trained gradient boosting tree prediction model for generating capacity prediction, and obtaining the theoretical generating capacity of the photovoltaic power generation system in a future period of time; acquiring the actual generating capacity of the photovoltaic power generation system in the corresponding future time period; the actual generating capacity and the theoretical generating capacity are compared and analyzed, and energy efficiency evaluation is completed; wherein on the basis of a countercurrent protection data rejecting mechanism, if the real-time monitoring data is in a countercurrent protection state, the related real-time monitoring data is rejected. According to the method, historical data, real-time monitoring data and a machine learning algorithm are combined, and accurate evaluation of the energy efficiency of the photovoltaic power generation system is realized through accurate power generation capacity prediction and a countercurrent protection data rejection mechanism.

Description

Technical Field

[0001] The invention relates to a photovoltaic power generation energy efficiency analysis method, device and storage medium applicable to a cigarette factory, and belongs to the technical field of photovoltaic power generation. Background Art

[0002] With the popularization of renewable energy, photovoltaic power generation is increasingly used in high-energy-consuming industrial fields such as cigarette factories. However, the instability of photovoltaic power generation and the lack of real-time and accurate energy efficiency analysis strategies have become key factors restricting its efficient use. These problems are particularly prominent in environments such as cigarette factories that have extremely high requirements for energy stability and continuity. In addition, due to system configuration errors or equipment failures, current reverse flow data will also be generated. If these data are not effectively eliminated, they will seriously affect the accuracy of energy efficiency analysis, leading to wrong decisions and energy efficiency optimization directions.

[0003] Therefore, it is particularly important to develop a response strategy that can monitor and accurately analyze the photovoltaic power generation efficiency in real time and effectively eliminate the reverse flow protection data. Summary of the invention

[0004] The purpose of the present invention is to provide a photovoltaic power generation energy efficiency analysis method, device and storage medium suitable for cigarette factories, aiming to accurately evaluate the energy efficiency of photovoltaic power generation systems, thereby providing strong data support for the system's optimized operation, maintenance decisions and performance improvements.

[0005] In order to achieve the above objectives / solve the above technical problems, the present invention is implemented by adopting the following technical solutions.

[0006] In one aspect, the present invention provides a photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory, comprising: Obtain historical power generation data and real-time monitoring data of photovoltaic power generation systems; Input the historical power generation data and real-time monitoring data into a pre-trained gradient boosting tree prediction model to perform power generation prediction, and obtain the theoretical power generation of the photovoltaic power generation system in the future; Obtain the actual power generation of the photovoltaic power generation system in the corresponding future time period; Compare and analyze the actual power generation with the theoretical power generation to complete energy efficiency evaluation; Among them, based on the backflow protection data elimination mechanism, if the real-time monitoring data is in the backflow protection state, the relevant real-time monitoring data is eliminated.

[0007] Optionally, the actual power generation is compared and analyzed with the theoretical power generation to complete energy efficiency evaluation, including: According to the actual power generation and the theoretical power generation, whether the power generation of the two is matched is compared. If they are matched, it is judged that the energy efficiency of the photovoltaic power generation system is normal; otherwise, it is judged whether the photovoltaic power generation system needs maintenance; If maintenance is required, the photovoltaic power generation system will be prompted to resolve the energy efficiency anomaly; otherwise, the temperature and humidity control and frequency conversion equipment in the cigarette factory will be controlled to dynamically adjust the output power to achieve energy-saving status.

[0008] Optionally, the photovoltaic power generation energy efficiency analysis method applicable to cigarette factories also includes data preprocessing, including: data cleaning of historical power generation data and real-time monitoring data, missing value filling, detection, identification and elimination of outliers, data standardization, data normalization, data encoding and data principal component analysis.

[0009] Optionally, the real-time monitoring data includes at least: current, voltage, temperature, light intensity and radiation.

[0010] Optionally, linear interpolation is performed on the missing values, which is expressed as: ; In the formula, represents the value of the data point after interpolation, represents the value of a known data point, Indicates the time point where interpolation is required. represents the timestamp of a known data point, The timestamp representing the next known data point.

[0011] Optionally, the outliers are detected, identified and removed based on the quartile range, including: The obtained first quartile and third quartile are calculated to obtain the quartile range, which is expressed as: ; In the formula, represents the interquartile range, represents the first quartile of the data set, Represents the third quartile of the data set; If the data point in the data set is less than or greater than , then the relevant data point is judged to be an abnormal value and the abnormal value is eliminated; otherwise, the relevant data point is judged to be a normal value and is retained.

[0012] Optionally, 25% of the data points in the data set are less than or equal to the first quartile, and 75% of the data points in the data set are less than or equal to the third quartile.

[0013] Optionally, the calculation formula corresponding to the principal component analysis of the data is as follows: ; In the formula, represents the data after dimensionality reduction, represents the transpose of the weight vector of the principal component, represents the original data matrix, Represents the mean vector of the data.

[0014] Optionally, the gradient boosting tree prediction model uses mean square error as its loss function, which is expressed as: ; In the formula, represents the loss function of the gradient boosting tree prediction model, Indicates the actual power generation. Represents the theoretical power generation predicted by the gradient boosting tree prediction model.

[0015] Optionally, the gradient boosting tree prediction model is a weighted sum of multiple weak learners, expressed as: ; In the formula, represents the number of weak learners, Indicates The weights of the weak learners, Indicates The prediction function of a weak learner, represents the initial prediction value.

[0016] Optionally, the The prediction function of a weak learner is expressed as: ; In the formula, Indicates The prediction function of the weak learner corresponding to the data point is represents the number of data points, Indicates The prediction value of a weak learner, Indicates The predicted value corresponding to each data point.

[0017] Optionally, the The weight of a weak learner is expressed as: ; In the formula, Indicates The reduced error rate of a weak learner.

[0018] Optionally, the photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory further includes: correcting the theoretical power generation predicted by the gradient boosting tree prediction model, expressed as: ; In the formula, represents the corrected predicted theoretical power generation, Indicates the correction amount calculated based on real-time data.

[0019] Optionally, based on the backflow protection data elimination mechanism, if the real-time monitoring data is in the backflow protection state, the relevant real-time monitoring data is eliminated, including: Acquiring the real-time monitoring data; Determine whether the timestamp corresponding to each data point in the real-time monitoring data is within the time period corresponding to the backflow event. If so, determine that the relevant data point is backflow data collected under the backflow protection state, and remove the backflow data; otherwise, retain it.

[0020] Optionally, the backflow event is represented by: ; In the formula, Indicates a backflow event, Indicates the start time of the backflow protection state. Indicates the end time of the backflow protection state.

[0021] In a second aspect, the present invention further provides a photovoltaic power generation energy efficiency analysis device applicable to a cigarette factory, comprising: A first data acquisition module is configured to acquire historical power generation data and real-time monitoring data of the photovoltaic power generation system; The backflow data elimination module is configured to eliminate the relevant real-time monitoring data if the real-time monitoring data is in a backflow protection state based on the backflow protection data elimination mechanism; The power generation prediction module is configured to input the historical power generation data and the real-time monitoring data into a pre-trained gradient boosting tree prediction model to perform power generation prediction, and obtain the theoretical power generation of the photovoltaic power generation system in the future period of time; The second data acquisition module is configured to acquire the actual power generation of the photovoltaic power generation system in a future time period; And, the energy efficiency evaluation module is configured to compare and analyze the actual power generation with the theoretical power generation to complete the energy efficiency evaluation.

[0022] In a third aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory as described in the first aspect.

[0023] Compared with the prior art, the present invention has the following beneficial effects: By introducing the gradient boosting tree prediction model, the present invention significantly improves the prediction accuracy of the theoretical power generation of the photovoltaic power generation system, and the system has the ability to monitor and compare the actual power generation with the theoretical power generation in real time, and can quickly identify the performance deviation of the photovoltaic power generation system, providing data support for timely adjustment; secondly, the present invention is also based on the reverse flow protection data elimination mechanism, which effectively eliminates the data interference during the reverse flow protection period to ensure the accuracy and reliability of energy efficiency analysis; in addition, after judging the energy efficiency degradation trend of the photovoltaic power generation system, the present invention also combines the frequency conversion control technology to dynamically adjust the output frequency to maintain the best energy-saving state, thereby optimizing the operating efficiency of the photovoltaic power generation system and realizing green and sustainable energy utilization. In general, the present invention significantly improves the prediction accuracy, data processing capability, maintenance efficiency and energy utilization efficiency of the photovoltaic power generation system by introducing innovations in multiple aspects such as machine learning algorithms, real-time data comparison, reverse flow protection data elimination, intelligent diagnosis and maintenance prompts, and frequency conversion control technology, providing a strong guarantee for the stable operation and efficient power generation of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 The figure shows a flow chart of an embodiment of the photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory according to the present invention. DETAILED DESCRIPTION

[0025] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. The embodiments of the present invention and the technical features in the embodiments may be combined with each other unless there is a conflict.

[0026] The term "and / or" is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the associated objects are in an "or" relationship.

[0027] Example 1 This embodiment introduces a photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory, which includes: Obtaining historical power generation data and real-time monitoring data of the photovoltaic power generation system; wherein the historical power generation data comes from the long-term operation record of the photovoltaic power generation system, including daily power generation, monthly power generation, etc., and is used to analyze the historical performance and trend of the photovoltaic power generation system; the real-time monitoring data is collected from the photovoltaic power generation system in real time, including but not limited to temperature, light intensity, radiation and other data, and is used to evaluate the operating status of the system in real time; Input the historical power generation data and real-time monitoring data into a pre-trained gradient boosting tree prediction model to perform power generation prediction, and obtain the theoretical power generation of the photovoltaic power generation system in the future; Obtain the actual power generation of the photovoltaic power generation system in the corresponding future time period. The actual power generation data can be obtained by direct measurement or reading system data; Compare and analyze the actual power generation with the theoretical power generation to complete energy efficiency evaluation; Among them, based on the backflow protection data elimination mechanism, if the real-time monitoring data is in the backflow protection state, the relevant real-time monitoring data is eliminated.

[0028] In the actual application process of this embodiment, it is first necessary to obtain the historical power generation data and real-time monitoring data of the photovoltaic power generation system. After the relevant data is obtained, the timestamp corresponding to each data point in the real-time monitoring data needs to be checked. If the timestamp is in the reverse flow protection state, the relevant data in this time period should be excluded. Next, the pre-trained gradient boosting tree prediction model is loaded, and the pre-processed historical power generation data and real-time monitoring data are used as inputs, and input into the gradient boosting tree prediction model for power generation prediction, and then the theoretical power generation prediction value of the photovoltaic power generation system in the future period is obtained, and the prediction value will be used for subsequent energy efficiency evaluation. Then, by directly measuring or reading the photovoltaic power generation system data, the actual power generation in the corresponding future time period is obtained, and the actual power generation is compared and analyzed with the theoretical power generation, and the error rate or deviation value is calculated. Finally, the energy efficiency of the photovoltaic power generation system is evaluated according to the error rate or deviation value. If the error rate or deviation value is large, it indicates that the system may have performance problems or need to be optimized.

[0029] In general, the photovoltaic power generation energy efficiency analysis method provided in this embodiment can accurately evaluate the energy efficiency status of the photovoltaic power generation system by obtaining historical power generation data and real-time monitoring data, performing power generation forecasts, obtaining actual power generation and performing energy efficiency evaluation, and provide strong support for system optimization and maintenance.

[0030] Example 2 refer to Figure 1Based on Example 1, this example will monitor and analyze the key parameters of the photovoltaic power generation system in real time, and combine the gradient boosting tree prediction model algorithm to conduct targeted analysis on whether the system has energy inefficiency problems. The specific implementation process is described as follows.

[0031] 1. Data Acquisition and Preprocessing The system is equipped with a variety of sensors for real-time collection of key parameters of the photovoltaic power generation system, such as current, voltage, temperature, light intensity, radiation, and historical power generation. Polyurethane insulation materials with a thickness of more than 10 cm are used to protect the sensors to reduce the impact of environmental factors on data collection accuracy.

[0032] After data collection, the relevant data needs to be preprocessed through the data preprocessing unit, including: data cleaning of historical power generation data and real-time monitoring data, missing value filling, detection, identification and elimination of outliers, data standardization, data normalization, data coding and data principal component analysis to ensure data quality and avoid analysis errors caused by data problems. The specific processing process is described as follows.

[0033] Data cleaning to remove invalid, erroneous or incomplete data records.

[0034] The missing values ​​are linearly interpolated according to the correlation of the data, which is expressed as: ; In the formula, represents the value of the data point after interpolation, represents the value of a known data point, Indicates the time point where interpolation is required. represents the timestamp of a known data point, The timestamp representing the next known data point.

[0035] The outliers are detected, identified and removed based on the interquartile range, including: The obtained first quartile and third quartile are calculated to obtain the quartile range, which is expressed as: ; In the formula, represents the interquartile range, which is used to define the boundaries of outliers. represents the first quartile of the data set, Represents the third quartile of the data set; If the data point in the data set is less than or greater than , then the relevant data point is judged to be an abnormal value and the abnormal value is eliminated; otherwise, the relevant data point is judged to be a normal value and is retained.

[0036] 25% of the data points in the data set are less than or equal to the first quartile, and 75% of the data points in the data set are less than or equal to the third quartile.

[0037] Scale the data proportionally to the specified range [-1,1] to achieve data standardization, expressed as: ; In the formula, represents the standardized value, represents the original data points, represents the average value, Represents standard deviation.

[0038] The normalization operation refers to scaling the data to a fixed range [0, 1]. This embodiment adopts the minimum-maximum normalization process, which is expressed as: ; In the formula, represents the normalized data, and Represents the minimum and maximum values ​​in the data set respectively.

[0039] Data encoding refers to data encoding using one-hot encoding, which is a method of converting categorical variables into a form that can be better processed by machine learning algorithms. For a categorical variable with 1 category, one-hot encoding creates binary columns, one for each category. If the data point belongs to the category, the value of the corresponding column is 1, otherwise it is 0. In this embodiment, Take 4.

[0040] Principal component analysis (PCA) is a dimensionality reduction technique used to reduce the number of features while retaining most of the variability in a dataset. The formula for PCA is as follows: ; In the formula, represents the data after dimensionality reduction, represents the transpose of the weight vector of the principal component, represents the original data matrix, Represents the mean vector of the data.

[0041] 2. Removal of upstream data The data during the backflow protection period may lead to inaccurate performance evaluation because it does not represent the normal operating status of the system. Analysis containing backflow data may affect the training of the gradient boosting tree prediction model, resulting in deviations in model predictions. Therefore, this embodiment is also designed with a backflow protection data elimination mechanism, which automatically identifies the data during the backflow protection period through preset logical conditions and flow control, and excludes erroneous data from the analysis data set to ensure the accuracy and reliability of the analysis results.

[0042] Reverse flow refers to the current flowing from the load to the power supply, which is opposite to the normal flow direction. Once reverse flow is detected, the system will record the timestamps of the start and end of the reverse flow and automatically exclude the data during the reverse flow protection period. The specific implementation process is described below.

[0043] Data collection and preprocessing: The system will collect key operating parameters of the photovoltaic power generation system in real time, including voltage, current and other data. All collected data will be stored with accurate timestamps for subsequent analysis and processing.

[0044] Reverse flow protection status detection: The system will monitor the direction and magnitude of the current in real time through sensors. Reverse flow means that the current flows to the grid, and normal flow means that the current flows to the load. The system detects the reverse flow protection status in real time. Once a reverse flow is detected, the system records the start and end time of the event. Through these preset logical conditions, the system can effectively detect the reverse flow protection status and take timely measures to ensure the safe and compliant operation of the photovoltaic power generation system. The specific implementation of these conditions helps to improve the reliability and stability of the system, and also provides important time markers for subsequent data analysis and fault diagnosis.

[0045] Data elimination: define the time period of the backflow protection state as a backflow event , expressed as: ; In the formula, Indicates the start time of the backflow protection state. Indicates the end time of the backflow protection state.

[0046] Furthermore, for each data point , the system will check its timestamp Is there a countercurrent event? If , then the data point The system will iterate over all data points and exclude those data points whose timestamps are within the time period of the reverse flow event.

[0047] 3. Construction and training of target model The gradient boosting tree prediction model in this embodiment is constructed based on historical power generation data and pre-processed real-time monitoring data, aiming to improve the accuracy of theoretical power generation prediction of photovoltaic power generation. The specific steps of target model construction are as follows: Definition of target model loss function: The gradient boosting tree prediction model uses mean square error (MSE) as its loss function, expressed as: ; In the formula, represents the loss function of the gradient boosting tree prediction model, Indicates the actual power generation. Represents the theoretical power generation predicted by the gradient boosting tree prediction model.

[0048] Target model framework: The gradient boosting tree prediction model is a weighted sum of multiple weak learners, expressed as: ; In the formula, represents the number of weak learners, Indicates The weights of the weak learners, Indicates The prediction function of a weak learner, represents the initial prediction value.

[0049] Target model training: including weak learner training and weight adjustment.

[0050] Weak learner training: Each weak learner The training is performed by minimizing the loss function corresponding to the weak prediction model. The prediction function of a weak learner is expressed as: ; In the formula, Indicates The prediction function of the weak learner corresponding to the data point is represents the number of data points, Indicates The prediction value of a weak learner, Indicates The predicted value corresponding to each data point.

[0051] Weight adjustment: The weight of each weak learner is calculated by the following formula, expressed as: ; In the formula, Indicates The weights of the weak learners, Indicates The reduced error rate of a weak learner.

[0052] In addition, this embodiment also involves the calculation of prediction residuals, and the relevant formula is as follows: ; In the formula, Indicates The residual of the data point, Indicates the actual value, Represents the predicted value.

[0053] In addition, this embodiment also updates the model by periodically using newly collected data to adapt to changes in environmental and operating conditions.

[0054] 4. Strategy Application and Energy Efficiency Analysis The preprocessed historical power generation data and real-time monitoring data are input into the pre-trained gradient boosting tree prediction model to predict power generation and obtain the theoretical power generation of the photovoltaic power generation system in the future.

[0055] It is worth noting that in order to ensure the actual application effect of the analysis system, this embodiment also performs feature selection and engineering on the relevant data. Feature selection refers to calculating feature correlation and selecting features with high correlation with power generation. Feature construction refers to creating time features (every hour) and historical average features.

[0056] Specifically, perform polynomial expansion on the features: ; In the formula, and represents the original features, Represents the expanded feature set.

[0057] And, create interaction terms between features: ; In addition, this embodiment further performs data correction on the theoretical power generation predicted by the gradient boosting tree prediction model, which is expressed as: ; In the formula, represents the corrected predicted theoretical power generation, Indicates the correction amount calculated based on real-time data.

[0058] Furthermore, the theoretical power generation predicted by the model is compared and analyzed with the actual power generation to complete the energy efficiency evaluation, including: comparing the actual power generation with the theoretical power generation to see whether the power generation of the two matches. If they match, it is determined that the energy efficiency of the photovoltaic power generation system is normal; otherwise, it is determined whether the photovoltaic power generation system needs maintenance; if maintenance is required, the photovoltaic power generation system is prompted to remove the energy efficiency anomaly; otherwise, the temperature and humidity control frequency conversion equipment of the cigarette factory is controlled to dynamically adjust the output power to achieve energy saving. The specific implementation process is described as follows.

[0059] Real-time monitoring and analysis: Real-time monitoring of actual power generation, and comparison and analysis with the theoretical power generation of the gradient boosting tree prediction model, through the added prediction correction algorithm to avoid the dead water zone effect, the system can monitor and analyze the difference between actual power generation and predicted power generation in real time, dynamically adjust the control strategy, reduce prediction errors, improve the adaptability and efficiency of the system, thereby avoiding the dead water zone effect, ensuring data flow and heat exchange efficiency. These algorithms work together to improve the stability and reliability of the photovoltaic power generation system and optimize energy utilization efficiency.

[0060] Intelligent diagnosis and maintenance prompts: The system automatically determines the energy efficiency degradation trend of the photovoltaic power generation system based on real-time comparison results and preset judgment strategies, thereby prompting staff to perform necessary maintenance or optimization.

[0061] Frequency conversion control: Based on the real-time comparison results and the preset judgment strategy, the system applies frequency conversion control technology to the temperature and humidity control frequency conversion equipment in the cigarette factory. By dynamically adjusting the output frequency, it maintains the best energy-saving state and optimizes the system operation efficiency.

[0062] Electric regulating valve control: In response to the 24-hour heat preservation and moisture retention process requirements of the cigarette factory, the system uses electric three-way regulating valves for temperature and humidity control equipment to ensure water supply to the terminal heat-using equipment side in the heat storage mode during the off-peak period, and adjusts the valve opening according to the heat demand.

[0063] In addition, this embodiment also optimizes the relevant control strategies. The system is based on various process parameters within the cycle, including temperature, humidity, and flow rate, which directly affect the performance and power generation of photovoltaic cells. Statistical analysis is performed on the process parameters within the cycle, and the average value and change trend are calculated to predict future energy demand and photovoltaic power generation efficiency. The set value of the cold water temperature of the refrigerator and the number of operating refrigeration units are determined. The cold water temperature directly affects the production environment of the cigarette factory. The appropriate temperature can improve production efficiency and product quality, and indirectly affect the stable operation of the photovoltaic power generation system. The water supply temperature heating strategy, loading strategy and load reduction strategy are used to achieve adaptive load adjustment of the refrigeration system.

[0064] In actual application, in response to the process requirements of 24-hour heat preservation and moisture retention in cigarette factories, the system summarizes the collected data according to the set time period (every hour), and calculates the average value and change trend of the process parameters (temperature, humidity). The system provides the temperature and humidity control frequency conversion device A of the cigarette factory based on the real-time comparison results and the preset judgment strategy. Device A maintains the best energy-saving state by dynamically adjusting the output frequency and optimizes the system operation efficiency. By maintaining a suitable production environment, reducing production interruptions and energy waste caused by environmental fluctuations, the stability and efficiency of the photovoltaic power generation system can be indirectly improved, thereby potentially increasing power generation.

[0065] Furthermore, the system determines the cold water temperature set value and the number of refrigeration units in operation, as well as the water supply temperature increase strategy, and implements intelligent loading and unloading strategies to respond to changes in load demand based on the average value and change trend of each process parameter within the cycle. When the load demand increases, the system intelligently increases the number of refrigeration units in operation or increases the operating frequency; when the load demand decreases, the system correspondingly reduces the number of units in operation or reduces the operating frequency. Intelligent loading and unloading strategies are implemented according to real-time load demand. The system will regularly evaluate and optimize the control strategy based on the latest process parameters and environmental conditions. Through the gradient boosting tree prediction model algorithm, it continuously learns and adapts to new operating conditions to achieve continuous performance improvement. By optimizing the operating parameters of the refrigeration system, reducing energy waste and improving energy utilization efficiency, the stability and efficiency of the photovoltaic power generation system are indirectly improved, which may increase power generation.

[0066] In summary, the photovoltaic power generation energy efficiency analysis strategy provided in this embodiment can effectively improve the energy efficiency management of photovoltaic power generation in cigarette factories, optimize energy consumption, reduce operation and maintenance costs, and improve system stability and reliability, which is at least reflected in the following aspects: Application of machine learning algorithms: By introducing the gradient boosting tree prediction model, this embodiment can more accurately predict the theoretical power generation. The machine learning algorithm can learn from historical data and optimize the prediction model. As time goes by, the accuracy of the prediction will continue to improve; Reverse flow protection data elimination: This embodiment designs a reverse flow protection data elimination mechanism, which can automatically exclude the data during the reverse flow protection period, ensuring the accuracy of energy efficiency analysis. This is particularly important for photovoltaic power generation systems, because reverse flow phenomena can affect power generation efficiency and system stability; Real-time data comparison: The system can monitor the actual power generation in real time and compare it with the predicted theoretical power generation. This real-time comparison mechanism can quickly identify performance deviations of the photovoltaic power generation system, so that timely measures can be taken to make adjustments; Intelligent diagnosis and maintenance prompts: The energy efficiency analysis system corresponding to this embodiment can automatically determine the energy efficiency degradation trend of the photovoltaic power generation system based on the real-time comparison results and the preset judgment strategy, and prompt the staff to perform necessary maintenance or optimization. This intelligent diagnosis function helps to detect problems in advance, reduce downtime, and improve the overall operation efficiency of the system; Frequency conversion control: Combined with frequency conversion control technology, the energy efficiency analysis system corresponding to this embodiment can dynamically adjust the output frequency, maintain the best energy-saving state, and optimize the system operation efficiency. The application of this technology helps to reduce energy waste and improve energy utilization efficiency; Electric regulating valve control: Electric three-way regulating valve is used to adjust the valve opening according to the heat demand to ensure water supply to the terminal heat-using equipment side in the heat storage mode during the off-peak period. This control method can more accurately meet the process requirements and reduce energy consumption.

[0067] Example 3 Based on the same inventive concept as that of Embodiment 1 or 2, this embodiment introduces a photovoltaic power generation energy efficiency analysis device applicable to a cigarette factory, which comprises: A first data acquisition module is configured to acquire historical power generation data and real-time monitoring data of the photovoltaic power generation system; The backflow data elimination module is configured to eliminate the relevant real-time monitoring data if the real-time monitoring data is in a backflow protection state based on the backflow protection data elimination mechanism; The power generation prediction module is configured to input the historical power generation data and the real-time monitoring data into a pre-trained gradient boosting tree prediction model to perform power generation prediction, and obtain the theoretical power generation of the photovoltaic power generation system in the future period of time; The second data acquisition module is configured to acquire the actual power generation of the photovoltaic power generation system in a corresponding future time period; And, the energy efficiency evaluation module is configured to compare and analyze the actual power generation with the theoretical power generation to complete the energy efficiency evaluation.

[0068] refer to Figure 1 The photovoltaic power generation energy efficiency analysis device applicable to cigarette factories can execute the photovoltaic power generation energy efficiency analysis method applicable to cigarette factories introduced in the above-mentioned embodiment 1 or 2, and the specific function implementation of each functional module is not repeated here.

[0069] Example 4 This embodiment introduces a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory as described in Embodiment 1 or 2 are implemented.

[0070] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0071] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0072] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0073] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0074] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory, characterized in that: include: Obtain historical power generation data and real-time monitoring data of photovoltaic power generation systems; Input the historical power generation data and real-time monitoring data into a pre-trained gradient boosting tree prediction model to perform power generation prediction, and obtain the theoretical power generation of the photovoltaic power generation system in the future; Obtain the actual power generation of the photovoltaic power generation system in the corresponding future time period; Compare and analyze the actual power generation with the theoretical power generation to complete energy efficiency evaluation; Among them, based on the backflow protection data elimination mechanism, if the real-time monitoring data is in the backflow protection state, the relevant real-time monitoring data is eliminated.

2. The photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory according to claim 1, characterized in that: Compare and analyze the actual power generation with the theoretical power generation to complete the energy efficiency evaluation, including: According to the actual power generation and the theoretical power generation, whether the power generation of the two is matched is compared. If they are matched, it is judged that the energy efficiency of the photovoltaic power generation system is normal; otherwise, it is judged whether the photovoltaic power generation system needs maintenance; If maintenance is required, the photovoltaic power generation system will be prompted to resolve the energy efficiency anomaly; otherwise, the temperature and humidity control and frequency conversion equipment in the cigarette factory will be controlled to dynamically adjust the output power to achieve energy-saving status.

3. The photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory according to claim 2 is characterized in that: It also includes data preprocessing, including: data cleaning, missing value filling, outlier detection, identification and elimination, data standardization, data normalization, data coding and data principal component analysis processing for historical power generation data and real-time monitoring data; The real-time monitoring data includes at least current, voltage, temperature, light intensity and radiation.

4. The photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory according to claim 3 is characterized in that: The missing values ​​are linearly interpolated, which is expressed as: ； In the formula, represents the value of the data point after interpolation, represents the value of a known data point, Indicates the time point where interpolation is required. represents the timestamp of a known data point, The timestamp representing the next known data point; The outliers are detected, identified and removed based on the quartile range, including: The obtained first quartile and third quartile are calculated to obtain the quartile range, which is expressed as: ； In the formula, represents the interquartile range, represents the first quartile of the data set, represents the third quartile of a data set; wherein 25% of the data points in the data set are less than or equal to the first quartile, and 75% of the data points in the data set are less than or equal to the third quartile; If the data point in the data set is less than or greater than , then the relevant data point is judged to be an abnormal value and the abnormal value is removed; otherwise, the relevant data point is judged to be a normal value and is retained; The calculation formula corresponding to the principal component analysis of the data is as follows: ； In the formula, represents the data after dimensionality reduction, represents the transpose of the weight vector of the principal component, represents the original data matrix, Represents the mean vector of the data.

5. The photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory according to claim 4 is characterized in that: The gradient boosting tree prediction model uses mean square error as its loss function, which is expressed as: ； In the formula, represents the loss function of the gradient boosting tree prediction model, Indicates the actual power generation. Represents the theoretical power generation predicted by the gradient boosting tree prediction model.

6. The photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory according to claim 5, characterized in that: The gradient boosting tree prediction model is a weighted sum of multiple weak learners, expressed as: ； In the formula, represents the number of weak learners, Indicates The weights of the weak learners, Indicates The prediction function of a weak learner, represents the initial prediction value; Among them, the The prediction function of a weak learner is expressed as: ； In the formula, Indicates The prediction function of the weak learner corresponding to the data point is represents the number of data points, Indicates The prediction value of a weak learner, Indicates The predicted value corresponding to the data point; The said The weight of a weak learner is expressed as: ； In the formula, Indicates The reduced error rate of a weak learner.

7. The photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory according to claim 6, characterized in that: It also includes correcting the theoretical power generation predicted by the gradient boosting tree prediction model, which is expressed as: ； In the formula, represents the corrected predicted theoretical power generation, Indicates the correction amount calculated based on real-time data.

8. The photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory according to claim 1, characterized in that: Based on the backflow protection data elimination mechanism, if the real-time monitoring data is in the backflow protection state, the relevant real-time monitoring data is eliminated, including: Acquiring the real-time monitoring data; Determine whether the timestamp corresponding to each data point in the real-time monitoring data is within the time period corresponding to the backflow event. If so, determine that the relevant data point is backflow data collected under the backflow protection state, and remove the backflow data; otherwise, retain it; Wherein, the countercurrent event is expressed as: ； In the formula, Indicates a backflow event, Indicates the start time of the backflow protection state. Indicates the end time of the backflow protection state.

9. A photovoltaic power generation energy efficiency analysis method and device suitable for a cigarette factory, characterized in that: include: A first data acquisition module is configured to acquire historical power generation data and real-time monitoring data of the photovoltaic power generation system; The backflow data elimination module is configured to eliminate the relevant real-time monitoring data if the real-time monitoring data is in a backflow protection state based on the backflow protection data elimination mechanism; The power generation prediction module is configured to input the historical power generation data and the real-time monitoring data into a pre-trained gradient boosting tree prediction model to perform power generation prediction, and obtain the theoretical power generation of the photovoltaic power generation system in the future period of time; The second data acquisition module is configured to acquire the actual power generation of the photovoltaic power generation system in a corresponding future time period; And, the energy efficiency evaluation module is configured to compare and analyze the actual power generation with the theoretical power generation to complete the energy efficiency evaluation.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the photovoltaic power generation energy efficiency analysis method applicable to a cigarette factory as described in any one of claims 1 to 8 are implemented.

Images