Optimized scheduling method for water-optical complementary system in combination with optical power prediction

Abstract

The invention provides a water-light complementary system optimization scheduling method combined with optical power prediction, and the method comprises the steps: collecting operation characteristic data, meteorological characteristic data and time characteristic data of a photovoltaic system, carrying out the preprocessing of the data, and constructing a multi-dimensional time series data set for training; constructing an optical power prediction model based on a long-short-term memory model, carrying out model training, extracting a nonlinear relation and time dependence characteristics between optical power and meteorological factors, and dynamically adjusting model parameters through an adaptive moment estimation optimization algorithm; the method comprises the following steps: constructing a scheduling optimization model of a water-light complementary system, configuring photovoltaic power generation installed capacity by adopting a non-dominated sorting genetic algorithm, optimizing scheduling to enable the photovoltaic power generation installed capacity to maximize power generation capacity and minimize light abandoning quantity when the photovoltaic power generation installed capacity meets constraint conditions, and adjusting water-power output according to the operation state of the system. And real-time scheduling control of water, electricity and photoelectricity is realized.

Description

Technical Field

[0001] This invention relates to the field of water-solar hybrid system optimization scheduling technology, and in particular to a water-solar hybrid system optimization scheduling method that combines optical power prediction. Background Technology

[0002] With the adjustment of the global energy structure and the continuous development of new power systems, photovoltaic power generation, as an energy mechanism that efficiently utilizes solar energy, exhibits intermittency and fluctuations due to weather changes and seasonal variations. Hydropower, on the other hand, has flexible regulation characteristics and a relatively stable power generation process. Hydropower-photovoltaic hybrid systems organically combine hydropower stations and photovoltaic power stations, leveraging the flexible regulation characteristics of hydropower to smooth out the fluctuations and intermittentities in photovoltaic output and improve energy utilization efficiency.

[0003] With the increasing openness of the electricity trading market, accurate ultra-short-term photovoltaic (PV) forecasts enable hydro-PV hybrid systems to better capitalize on spot market price fluctuations, thus improving power system stability and facilitating optimized scheduling of these systems. When participating in spot trading, accurate PV forecasts allow for the rational submission of power generation plans, preventing default risks due to inaccurate capacity estimates, insufficient or excessive power generation, and reducing penalties for such defaults. When photovoltaic power forecasts indicate high PV output and high market prices in the near future, the system can plan ahead, increasing combined PV and hydropower output during that period to sell more electricity to the market and generate higher revenue; conversely, it can appropriately reduce power generation to lower costs.

[0004] However, traditional forecasting methods, such as time series analysis and physical modeling, rely primarily on the time-series characteristics of historical data. These methods struggle to accurately capture the highly nonlinear relationship between complex and ever-changing influencing factors and optical power, and are unable to handle the rapid dynamic changes in meteorological data under extreme weather conditions. This results in low forecast accuracy, making it difficult to meet the stringent requirements of real-time and precise scheduling for hydro-solar hybrid systems. Furthermore, traditional hydro-solar hybrid systems only consider some constraints, leading to low overall energy utilization efficiency and severely limiting the overall operational efficiency and reliability of the system.

[0005] In recent years, with the rapid development of artificial intelligence technology, deep learning technology has been widely used in the field of optical power prediction. Among the existing technologies, prediction algorithms based on Convolutional Neural Network (CNN), Support Vector Machine (SVM), and Extreme Learning Machine (BPNN) have improved the prediction accuracy to a certain extent, but still have many shortcomings: (1) When processing time series data such as optical power, the ability to capture long-term time dependence is limited, and the ability to remember information over a long period of time is poor; (2) There is a lack of ability to fuse and analyze multi-source data, and it is difficult to comprehensively consider meteorological, historical data, and operational status data to improve prediction accuracy; (3) The generalization ability is poor, and it is difficult to adapt to the diversity of optical power data under different regions and different climatic conditions. Summary of the Invention

[0006] The present invention aims to at least partially solve one of the technical problems in the related art.

[0007] Therefore, the first objective of this invention is to propose an optimized scheduling method for a water-solar complementary system that incorporates optical power prediction.

[0008] The second objective of this invention is to propose an optimized scheduling device for a water-solar hybrid system that incorporates optical power prediction.

[0009] The third objective of this invention is to provide an electronic device.

[0010] The fourth objective of this invention is to provide a computer-readable storage medium.

[0011] The fifth objective of this invention is to provide a computer program product.

[0012] To achieve the above objectives, a first aspect of the present invention proposes an optimized scheduling method for a water-solar hybrid system that incorporates optical power prediction, comprising:

[0013] Collect operational characteristic data, meteorological characteristic data, and temporal characteristic data of photovoltaic systems, and preprocess the data, including missing value reconstruction, smoothing of non-stationary components, standardization or normalization, and converting categorical data into numerical codes to construct a multidimensional time series dataset for training.

[0014] A light power prediction model based on a long short-term memory model is constructed. The model is trained using the multi-dimensional time series dataset. The nonlinear relationship and time-dependent characteristics between light power and meteorological factors are extracted. The model parameters are dynamically adjusted by an adaptive moment estimation optimization algorithm. The root mean square error and mean absolute error are used as loss functions for model optimization.

[0015] The system collects real-time data on water level, flow rate, and power generation of hydropower stations, as well as solar power stations on light intensity, temperature, relative humidity, and power generation, and transmits the data to the dispatch center.

[0016] A scheduling optimization model for a hydro-solar hybrid system is constructed. A non-dominated sorting genetic algorithm is used to configure the photovoltaic power generation capacity. The optimized scheduling maximizes the power generation and minimizes the curtailment of photovoltaic power generation while meeting the constraints. The hydropower output is adjusted according to the system operating status to achieve real-time scheduling and control of hydropower and photovoltaic power.

[0017] To achieve the above objectives, a second aspect of the present invention provides an optimized scheduling device for a water-solar hybrid system incorporating optical power prediction, comprising:

[0018] The photovoltaic output data acquisition module is used to collect the operating characteristic data, meteorological characteristic data and time characteristic data of the photovoltaic system, and to preprocess the data, including missing value reconstruction, smoothing of non-stationary components, standardization or normalization, and conversion of categorical data into numerical codes to construct a multidimensional time series dataset for training.

[0019] The prediction module is used to construct a light power prediction model based on a long short-term memory model. The model is trained using the multi-dimensional time series dataset, and the nonlinear relationship and time-dependent features between light power and meteorological factors are extracted. The model parameters are dynamically adjusted through an adaptive moment estimation optimization algorithm, and the root mean square error and mean absolute error are used as loss functions for model optimization.

[0020] The hydropower station data acquisition module is used to collect real-time data on water level, flow rate, and power generation of the hydropower station, as well as the light intensity, temperature, relative humidity, and power generation of the photovoltaic power station, and transmit the data to the dispatch center.

[0021] The execution scheduling optimization module is used to construct a scheduling optimization model for the hydro-solar hybrid system. It uses a non-dominated sorting genetic algorithm to configure the photovoltaic power generation capacity, optimizes the scheduling so that the photovoltaic installed capacity maximizes power generation and minimizes curtailment under the constraints. It also adjusts the hydropower output according to the system operating status to achieve real-time scheduling and control of hydropower and photovoltaic power.

[0022] To achieve the above objectives, a third aspect of the present invention provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor;

[0023] The memory stores computer-executed instructions;

[0024] The processor executes computer execution instructions stored in the memory to implement the method as described in any one of the first aspects.

[0025] To achieve the above objectives, a fourth aspect of the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of the first aspects.

[0026] To achieve the above objectives, a fifth aspect of the present invention provides a computer program product that, when executed by a processor, implements the method described in any one of the first aspects.

[0027] The technical solutions provided by the embodiments of the present invention bring at least the following beneficial effects:

[0028] (1) Data fusion: This invention combines real-time data and historical data, focusing on processing photovoltaic system operation status data and historical sunshine data and meteorological data in time series form. By using LSTM with its gating mechanism, it can effectively remember and forget information at different time scales, accurately capture the trend of changes in solar power over hours to days, especially the fluctuation pattern of solar power under the influence of factors such as sudden weather changes and day-night cycles in the ultra-short term (future 15 min-4 h).

[0029] (2) Dynamic Optimization using Adaptive Moment Estimation (Adam) Algorithm: This invention employs the Adaptive Moment Estimation (Adam) algorithm for dynamic optimization. The learning rate is adaptively adjusted based on the gradients of different parameters, which not only accelerates model convergence but also avoids getting trapped in local optima. Using root mean square error (RMSE) and mean absolute error (MAE) as loss functions, the model parameters are continuously optimized by minimizing the loss function, significantly improving the accuracy of optical power prediction.

[0030] (3) Multi-objective optimization technology: This invention adopts multi-objective optimization technology with the goal of maximizing power generation and minimizing load shortage rate. Under different seasons, different weather conditions and different power grid load demands, this technology can quickly and accurately select the optimal scheduling scheme that fits the current operating conditions from the Pareto optimal solution set.

[0031] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0032] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0033] Figure 1 This is a flowchart illustrating an optimized scheduling method for a water-solar hybrid system that incorporates optical power prediction, provided in an embodiment of the present invention.

[0034] Figure 2 This is a flowchart illustrating an optimized scheduling method for a water-solar hybrid system that incorporates optical power prediction, provided in an embodiment of the present invention.

[0035] Figure 3 This is a schematic diagram of the LSTM process provided in an embodiment of the present invention;

[0036] Figure 4 This is a schematic diagram of the structure of an optimized scheduling device for a water-solar hybrid system that combines optical power prediction, provided in an embodiment of the present invention. Detailed Implementation

[0037] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0038] Figure 1 and Figure 2 This is a flowchart illustrating an optimized scheduling method for a water-solar complementary system combining optical power prediction, provided by an embodiment of the present invention. This application utilizes ultra-short-term optical power prediction to achieve optimized scheduling of the water-solar complementary system. The system mainly comprises three parts: ultra-short-term optical power prediction, data collection and transmission, and water-solar complementary system scheduling optimization.

[0039] (1) Photovoltaic ultra-short-term prediction part: This part is responsible for the ultra-short-term prediction of photovoltaic power. This includes collecting historical data of photovoltaic system output, meteorological data and operating status data, preprocessing and extracting time features; constructing a long short-term memory model to achieve ultra-short-term prediction training, presenting the prediction results in the form of time series, evaluating the prediction results and optimizing the prediction model.

[0040] (2) Data acquisition and transmission section: This section collects data such as water level, flow rate, and power generation of hydropower stations in real time, as well as data such as light intensity, temperature, relative humidity, and power generation of photovoltaic power stations, and transmits the data to the dispatch center.

[0041] (3) Optimized Scheduling of the Hydro-Solar Complementary System: This part realizes the optimized scheduling of the hydro-solar complementary system. It includes constructing a scheduling optimization model for the hydro-solar complementary system, generating specific scheduling decision instructions based on the optimal solution obtained by the multi-objective optimization module, and specifically executing control operations on the hydropower station and photovoltaic power station; continuously monitoring the overall operating status of the hydro-solar complementary system and feeding it back to the multi-objective optimization module and the scheduling decision generation module so as to adjust and optimize the scheduling scheme and ensure that the system is always in the optimal operating state.

[0042] Specifically, such as Figure 1and Figure 2 As shown, the method includes the following steps:

[0043] Step S1: Collect the operating characteristic data, meteorological characteristic data and time characteristic data of the photovoltaic system, and preprocess the data, including missing value reconstruction, smoothing of non-stationary components, standardization or normalization, and convert categorical data into numerical codes to construct a multidimensional time series dataset for training.

[0044] In the implementation of this application, it is first necessary to collect operational characteristic data, meteorological characteristic data, and temporal characteristic data of the photovoltaic system. The comprehensiveness and accuracy of the data are crucial for the subsequent training of the model. Therefore, we collect these data in three main categories: operational characteristic data of the photovoltaic system, meteorological characteristic data, and temporal characteristic data.

[0045] The main operational characteristic data of photovoltaic systems include the following:

[0046] Historical output data of ultra-short-term photovoltaic power: These data record the actual photovoltaic power output of the photovoltaic system at a specific moment.

[0047] Equipment operation status data: This type of data includes the operating status of each device in the photovoltaic system, such as whether the device is in normal operation, whether a fault has occurred, and equipment maintenance records.

[0048] Historical energy production and consumption data: This includes historical energy production and consumption data for photovoltaic systems, reflecting the system's energy efficiency and load status.

[0049] Meteorological characteristic data include:

[0050] Meteorological conditions such as temperature, humidity, air pressure, cloud thickness, and cloud movement speed directly affect the power generation efficiency of photovoltaic systems.

[0051] Sunlight intensity: Sunlight intensity is the main driving force for photovoltaic systems to generate electricity, and the accuracy of the data is crucial for power generation prediction.

[0052] Time-related data include:

[0053] The time information collected during data collection, such as hour, date, day of the week, and season, has a significant impact on the fluctuations and trends in power generation.

[0054] After data collection is complete, the next step is to preprocess the data to ensure its quality and usability. Preprocessing includes:

[0055] Missing value reconstruction: During data collection, data may be missing. Therefore, appropriate interpolation methods or mean imputation are used to reconstruct the missing data, ensuring the integrity of the dataset and avoiding bias in model training due to missing data.

[0056] Non-stationary component smoothing: Since the output data of photovoltaic systems typically exhibits some volatility, especially with potential for drastic changes in the short term, smoothing is necessary to remove short-term fluctuations or noise and extract the long-term trend. Common techniques used in this step include moving averages and exponentially weighted averages.

[0057] Standardization and Normalization: To eliminate differences between different feature units, data needs to be standardized or normalized. Standardization refers to transforming data according to its mean and standard deviation to make the data conform to a standard normal distribution; while normalization compresses the data to a specific range (usually between 0 and 1) to eliminate the influence of different features on model training.

[0058] The data normalization formula is:

[0059]

[0060] In the formula X g The normalized value; X i X a Let X be the minimum and maximum values ​​of the initial sequence, respectively, and let X be the initial value of the sequence.

[0061] Categorical data encoding: For non-numerical categorical data, such as equipment status or seasonal information, encoding is required. Common methods include label encoding and one-hot encoding, the purpose of which is to convert categorical data into numerical form for processing by machine learning models.

[0062] Historical data alignment: To ensure the consistency of time-series data, all data needs to be aligned at fixed time intervals. For example, data from the 72 hours prior to the test date can be aligned at 15-minute intervals, so that the data at each time point has the same time granularity, providing a unified standard for subsequent analysis.

[0063] Finally, the data is divided into training and test sets according to a fixed ratio, typically 8:2 or 7:3. This approach ensures diversity in the model during training while allowing for accurate evaluation on the test set.

[0064] Correlation analysis is a crucial step in data preprocessing. By calculating the correlation between various data features, the impact of different features on photovoltaic system performance can be assessed. Features with higher correlations indicate a stronger linear relationship, which helps in selecting the most predictive features. The correlation analysis formula is as follows:

[0065]

[0066] In the formula, R is the correlation coefficient between the two variables; d i 2 R represents the positional difference between two variables after sorting them from smallest to largest in the original data; n is the size of the original dataset; the larger R is, the higher the correlation between the variables.

[0067] After these data preprocessing steps, the present invention can construct a multidimensional time series dataset, which is suitable for training a prediction model for photovoltaic system power generation, providing a scientific basis for the performance optimization and operation prediction of photovoltaic systems.

[0068] Step S2: Construct a light power prediction model based on a long short-term memory model, train the model using the multidimensional time series dataset, extract the nonlinear relationship and time-dependent features between light power and meteorological factors, and dynamically adjust the model parameters using an adaptive moment estimation optimization algorithm, with root mean square error and mean absolute error as loss functions for model optimization.

[0069] In this embodiment, the first step is to construct an optical power prediction model based on a Long Short-Term Memory (LSTM) network. (Refer to...) Figure 3 This model utilizes the multidimensional time-series dataset collected and processed in step S1, including historical power output data, meteorological data, and operational status data of the photovoltaic system. During the construction process, the model can automatically extract the nonlinear relationship and time-dependent characteristics between photovoltaic power and meteorological factors from this data, providing accurate predictions of the ultra-short-term photovoltaic power output.

[0070] First, by utilizing historical data, the LSTM model can identify the complex relationship between photovoltaic power output and meteorological factors (such as temperature, humidity, and light intensity). Since changes in photovoltaic power output are not only related to meteorological conditions but also influenced by time factors, the LSTM model is particularly suitable for capturing the time-dependent characteristics in such time-series data. In this way, the model can identify the patterns of photovoltaic system power variation in a short period and make accurate ultra-short-term predictions.

[0071] During training, the model was optimized using an adaptive moment estimation algorithm, which dynamically adjusts the model's parameters to better fit the training data. To ensure the optimization effect, the root mean square error (RMSE) and mean absolute error (MAE) were selected as the loss functions. These metrics effectively reflect the difference between the model's predicted values ​​and the actual values, thereby helping to optimize the model's performance.

[0072] Specifically, the statistical error E S for:

[0073]

[0074] The mean absolute error (MAE) is:

[0075]

[0076] The root mean square error (RMSE) is:

[0077]

[0078] Accuracy C Rnew for:

[0079]

[0080] The correlation coefficient r is:

[0081]

[0082] In the formula, n is the number of samples, and P Pt Let P be the actual photovoltaic power at time t. Mt Let C be the predicted photovoltaic power at time t. t Let be the power-on capacity at time t; This represents the average actual power during the error statistics period. This represents the average predicted power over the error statistics period.

[0083] After training, the model's prediction results need to be evaluated. These error evaluation metrics allow for a comprehensive and objective assessment of the model's predictive ability. Finally, by plotting a comparison between the predicted and actual results, the model's performance in ultra-short-term optical power prediction can be visually demonstrated. In this way, users can clearly see the model's actual prediction performance, further validating its feasibility and application value.

[0084] In summary, the LSTM-based photovoltaic power prediction model, trained using historical data, meteorological data, and temporal characteristics of the photovoltaic system, can accurately capture the complex relationship between photovoltaic power and meteorological factors, and its performance can be evaluated through various error metrics. This method provides strong technical support for ultra-short-term photovoltaic power prediction and has broad application prospects.

[0085] Step S3: Real-time data collection of water level, flow rate, and power generation of hydropower stations, as well as light intensity, temperature, relative humidity, and power generation of photovoltaic power stations, and transmission of the data to the dispatch center.

[0086] In this embodiment, step S3 involves real-time collection and transmission of key operational data from hydropower stations and photovoltaic power stations to the dispatch center. The main task of this step is to ensure the timely and accurate transmission of various types of data, and to guarantee the security and integrity of the data through standardized data formats, so that the dispatch center can make rapid responses and optimization adjustments.

[0087] First, the system needs to collect various operational data from hydropower stations and photovoltaic power stations in real time. Specifically, the operational data for hydropower stations includes water level, flow rate, and power generation. This data reflects the operating status and power generation capacity of the hydropower station, which is crucial for the dispatch center. It helps to understand the energy production status of the hydropower station in a timely manner and adjust the reservoir water level to ensure maximum power generation efficiency. The data for photovoltaic power stations includes solar irradiance, temperature, relative humidity, and power generation. The power generation capacity of photovoltaic power stations is affected by these meteorological conditions, so it is necessary to acquire this data in real time to assess power generation capacity and predict future output.

[0088] To ensure data real-time performance, the system needs to set a reasonable data acquisition frequency and automatically collect this data within specified time intervals. For example, water level and flow data of a hydropower station may be collected once per hour, while environmental parameters of a photovoltaic power station (such as light intensity, temperature, humidity, etc.) can be collected at a higher frequency (such as once every 15 minutes) depending on the frequency of weather changes.

[0089] The collected data will be encoded using a standard format for stable and efficient transmission. The use of a standard format ensures that data from different sources can be processed and analyzed uniformly, avoiding parsing errors caused by inconsistent data formats. Furthermore, to guarantee data security and integrity, the system will employ encryption algorithms to protect the data and prevent tampering or loss during transmission.

[0090] All collected data will be transmitted to the dispatch center within a set time. The dispatch center will monitor the received data in real time and make optimizations as needed. For example, when the water level at a hydropower station is low, the dispatch center may optimize power generation scheduling based on flow data; or when the sunlight intensity at a photovoltaic power station is low, the dispatch center may adjust power dispatch in advance to ensure the stability of the power grid.

[0091] Through this process, the dispatch center can obtain real-time operating data from hydropower stations and photovoltaic power stations, thereby achieving precise energy dispatch and optimization, and improving the operating efficiency and reliability of the power system.

[0092] Step S4: Construct a scheduling optimization model for the hydro-solar hybrid system. Use a non-dominated sorting genetic algorithm to configure the photovoltaic power generation capacity. Optimize the scheduling so that the photovoltaic installed capacity maximizes power generation and minimizes curtailment under the constraints. Adjust the hydropower output according to the system operating status to achieve real-time scheduling and control of hydropower and photovoltaic power.

[0093] In this embodiment, step S4 involves constructing a hydro-solar hybrid system scheduling optimization model, employing a non-dominated sorting genetic algorithm (NSGA-II) to optimize the configuration of photovoltaic power generation capacity, and realizing real-time scheduling control of hydropower and photovoltaic power. The goal of this model is to maximize photovoltaic power generation, minimize curtailment, and adjust hydropower output according to the system operating status to achieve the optimal scheduling effect of hydro-solar hybridization.

[0094] First, the objective function of the water-solar hybrid system scheduling optimization model is set as the optimization of multiple indicators. Specifically, the objective function includes the following aspects:

[0095] Maximizing power generation: This objective aims to maximize the system's power generation by optimizing the output configuration of photovoltaic and hydroelectric generators. The objective function is:

[0096]

[0097] In the formula P Pt P ht,i These represent the photovoltaic power output and the output of the i-th hydroelectric generator unit at time t, respectively. T is the total number of intervals within the time range, m is the number of hydroelectric generator units, and Δt is the time interval.

[0098] Minimize Load Deficiency Rate (LPSP): This objective aims to ensure that the system's generating capacity meets load demand and avoids power shortages. The formula for the load deficiency rate is:

[0099]

[0100] Another relevant indicator is power fluctuation, which is calculated using the following formula:

[0101]

[0102] In the formula f LPSP For load power shortage rate, γ L For the power fluctuation between the output and load of the hydro-solar hybrid system, P Lt Let t be the load at time t.

[0103] Minimize curtailment: Curtailment refers to the amount of photovoltaic power that cannot be effectively utilized due to overcapacity in certain situations. The objective function for minimizing curtailment is:

[0104]

[0105] In the formula This represents the amount of light discarded at time t.

[0106] To ensure that the model operates in accordance with physical and engineering realities, the following constraints must be considered:

[0107] V i,min ≤V i ≤V i,max

[0108] P h,i,min ≤P j,i ≤P j,i,max

[0109] 0≤P Pt ≤P max

[0110] 0≤P Pt ≤P P,max

[0111] 0≤N pv ≤N pv,max

[0112] P pt ≤λP L ≤P wp

[0113] P f(t) =P pt +P ht,i +Q Pt

[0114] In the above formula, V i V represents the reservoir capacity of the i-th adjustable hydropower station. i,min V i,max Let P represent the minimum and maximum reservoir capacities of the i-th adjustable hydropower station, respectively. wp P represents the reserve constraint caused by the integration of a multi-energy system into the power grid. max P represents the maximum acceptable photovoltaic capacity of the power grid.h,i Let P represent the output of the i-th adjustable hydroelectric power station. h,i,min P h,i,max Let P represent the minimum and maximum output of the i-th adjustable hydropower station, respectively. When the adjustable hydropower station is a horizontal pumped storage power station, P P,max N represents the maximum output of the photovoltaic power station in a multi-energy system. pv N represents the installed capacity of a multi-energy photovoltaic power station. pv,max P represents the maximum allowable installed capacity of a multi-energy photovoltaic power station. f(t) Q represents the load of the multi-energy system at time t. Pt This represents the amount of wind and solar power curtailment in the multi-energy system at time t.

[0115] By optimizing the scheduling relationship between photovoltaic (PV) and hydropower outputs, and combining the above objective function and constraints, the model can maximize PV power generation, minimize curtailment, and achieve rational hydropower scheduling while meeting load demand. This optimized scheduling model can effectively improve the overall power generation efficiency of the hydro-PV hybrid system and provide scientific decision support for practical energy dispatching.

[0116] To achieve the above embodiments, the present invention also proposes an optimized scheduling device for a water-solar complementary system that combines optical power prediction. Figure 4 This is a schematic diagram of a water-solar hybrid system optimization scheduling device that incorporates optical power prediction, provided as an embodiment of the present invention. Figure 4 As shown, the device includes:

[0117] The photovoltaic output data acquisition module is used to collect the operating characteristic data, meteorological characteristic data and time characteristic data of the photovoltaic system, and to preprocess the data, including missing value reconstruction, smoothing of non-stationary components, standardization or normalization, and conversion of categorical data into numerical codes to construct a multidimensional time series dataset for training.

[0118] The prediction module is used to construct a light power prediction model based on a long short-term memory model. The model is trained using the multi-dimensional time series dataset, and the nonlinear relationship and time-dependent features between light power and meteorological factors are extracted. The model parameters are dynamically adjusted through an adaptive moment estimation optimization algorithm, and the root mean square error and mean absolute error are used as loss functions for model optimization.

[0119] The hydropower station data acquisition module is used to collect real-time data on water level, flow rate, and power generation of the hydropower station, as well as the light intensity, temperature, relative humidity, and power generation of the photovoltaic power station, and transmit the data to the dispatch center.

[0120] The execution scheduling optimization module is used to construct a scheduling optimization model for the hydro-solar hybrid system. It uses a non-dominated sorting genetic algorithm to configure the photovoltaic power generation capacity, optimizes the scheduling so that the photovoltaic installed capacity maximizes power generation and minimizes curtailment under the constraints. It also adjusts the hydropower output according to the system operating status to achieve real-time scheduling and control of hydropower and photovoltaic power.

[0121] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0122] To implement the above embodiments, the present invention also proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer execution instructions; the processor executes the computer execution instructions stored in the memory to implement the method provided in the foregoing embodiments.

[0123] To implement the above embodiments, the present invention also proposes a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods provided in the foregoing embodiments.

[0124] To implement the above embodiments, the present invention also proposes a computer program product, including a computer program that, when executed by a processor, implements the methods provided in the foregoing embodiments.

[0125] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in this invention all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0126] It should be noted that personal information collected from users should be used for legitimate and reasonable purposes and should not be shared or sold outside of these legitimate uses. Furthermore, such collection / sharing should only be conducted after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign an agreement / authorization that includes authorization of relevant user information before the user uses the function. In addition, any necessary steps must be taken to protect and safeguard access to such personal information data and ensure that others with access to personal information data comply with their privacy policies and procedures.

[0127] This invention is intended to provide implementation schemes for users to selectively prevent the use or access to personal information data. That is, this disclosure is intended to provide hardware and / or software to prevent or block access to such personal information data. Once personal information data is no longer needed, risks can be minimized by restricting data collection and deleting data. Furthermore, where applicable, such personal information can be de-identified to protect user privacy.

[0128] In the foregoing descriptions of the embodiments, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0129] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0130] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of the invention pertain.

[0131] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0132] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0133] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it includes one or a combination of the steps of the method embodiments.

[0134] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0135] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

[0136] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0137] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for optimizing the scheduling of a water-solar hybrid system by combining optical power prediction, characterized in that, Includes the following steps: Collect operational characteristic data, meteorological characteristic data, and temporal characteristic data of photovoltaic systems, and preprocess the data, including missing value reconstruction, smoothing of non-stationary components, standardization or normalization, and converting categorical data into numerical codes to construct a multidimensional time series dataset for training. A light power prediction model based on a long short-term memory model is constructed. The model is trained using the multi-dimensional time series dataset. The nonlinear relationship and time-dependent characteristics between light power and meteorological factors are extracted. The model parameters are dynamically adjusted by an adaptive moment estimation optimization algorithm. The root mean square error and mean absolute error are used as loss functions for model optimization. The system collects real-time data on water level, flow rate, and power generation of hydropower stations, as well as solar power stations on light intensity, temperature, relative humidity, and power generation, and transmits the data to the dispatch center. A scheduling optimization model for a hydro-solar hybrid system is constructed. A non-dominated sorting genetic algorithm is used to configure the photovoltaic power generation capacity. The optimized scheduling maximizes the power generation and minimizes the curtailment of photovoltaic power generation while meeting the constraints. The hydropower output is adjusted according to the system operating status to achieve real-time scheduling and control of hydropower and photovoltaic power.

2. The method according to claim 1, characterized in that, The process involves collecting operational characteristic data, meteorological characteristic data, and temporal characteristic data from the photovoltaic system, and preprocessing the data, including missing value reconstruction, smoothing of non-stationary components, standardization or normalization, and converting categorical data into numerical codes to construct a multidimensional time series dataset for training, including: Collect historical data, meteorological data, and operational status of the photovoltaic system. Historical data includes historical ultra-short-term photovoltaic power output data; historical output data includes operational characteristic data, meteorological characteristic data, and temporal characteristic data; operational characteristic data includes irradiance, actual photovoltaic power output, photovoltaic equipment operational status, equipment maintenance records, and historical energy production and consumption data; meteorological characteristic data includes temperature, relative humidity, air pressure, cloud thickness and movement speed, and irradiance; temporal characteristic data includes the hour, day, week, and corresponding season at the time of data collection. The collected data were preprocessed, missing values ​​were reconstructed, the original data were expanded and enhanced, the peak values ​​and non-stationary components in the historical data set were smoothed, and the data were divided into training set and test set according to a fixed ratio; numerical data were standardized or normalized, and categorical data were converted into numerical codes; the historical data of the photovoltaic system operation for nearly 72 hours on the day to be tested were aligned at 15-minute intervals. The data normalization formula is as follows: In the formula X g The normalized value; X i X a Let X be the minimum and maximum values ​​of the initial sequence, respectively, and let X be the initial value of the sequence. The formula for data correlation is: In the formula, R is the correlation coefficient between the two variables; d i 2 R represents the positional difference between two variables after sorting them from smallest to largest in the original data; n is the size of the original dataset; the larger R is, the higher the correlation between the variables.

3. The method according to claim 2, characterized in that, The construction of the optical power prediction model based on the long short-term memory model involves training the model using the multi-dimensional time series dataset, extracting the nonlinear relationship and time-dependent features between optical power and meteorological factors, and dynamically adjusting the model parameters using an adaptive moment estimation optimization algorithm. Model optimization is performed using root mean square error and mean absolute error as loss functions, including: A deep learning-based ultra-short-term photovoltaic power prediction model is constructed. The training set data is input into the prediction model to train and optimize the model. Temporal features and nonlinear relationships between photovoltaic power and meteorological factors are extracted from the collected historical data, meteorological data, and operational status data of the photovoltaic system output. The extracted temporal features are fused with the original data, and ultra-short-term features are integrated. The processed data is then input again into the LSTM-based ultra-short-term photovoltaic power prediction model to predict photovoltaic output. The prediction results are evaluated using an error evaluation index, which includes systematic error E. S Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Accuracy C Rnew The correlation coefficient r is calculated, and a comparison chart of the predicted results and the actual results is plotted to intuitively demonstrate the predictive ability of the model. Wherein, the systematic error E S for: The mean absolute error (MAE) is: The root mean square error (RMSE) is: Accuracy C Rnew for: The correlation coefficient r is: In the formula, n is the number of samples, and P Pt Let P be the actual photovoltaic power at time t. Mt Let C be the predicted photovoltaic power at time t. t Let be the power-on capacity at time t; This represents the average actual power during the error statistics period. This represents the average predicted power over the error statistics period.

4. The method according to claim 3, characterized in that, The system collects real-time data on water level, flow rate, and power generation of hydropower stations, as well as solar irradiance, temperature, relative humidity, and power generation of photovoltaic power stations, and transmits the data to the dispatch center, including: Real-time data collection of water level, flow rate, and power generation of hydropower stations; and solar power station data of irradiance, temperature, relative humidity, and power generation. Set the frequency of data collection and transmission, encode the data in a standard format, and transmit it to the dispatch center to ensure data security and integrity, so as to make timely adjustments and optimizations.

5. The method according to claim 4, characterized in that, The construction of the water-solar hybrid system scheduling optimization model includes: The objective function of the water-solar hybrid system scheduling optimization model is set as follows: Maximum power generation: In the formula P Pt P ht,i These are the photovoltaic power output and the output of the i-th hydroelectric generator unit at time t, respectively. T is the total number of intervals within the time range, m is the number of hydroelectric generator units, and Δt is the time interval. Minimum load shortage rate: In the formula f LPSP For load power shortage rate, γ L For the power fluctuation between the output and load of the hydro-solar hybrid system, P Lt The load at time t; Minimum light wastage: In the formula This represents the amount of light discarded at time t. Constraints: In i,min ≤V i ≤V i,max P h,i,min ≤P h,i ≤P h,i,max 0≤P Pt ≤P max 0≤P Pt ≤P P,max 0≤N pv ≤N pv,max P pt ≤λP L ≤P wp P f(t) =P pt +P ht,i +Q Pt In the above formula, V i V represents the reservoir capacity of the i-th adjustable hydropower station. i,min V i,max Let P represent the minimum and maximum reservoir capacities of the i-th adjustable hydropower station, respectively. wp P represents the reserve constraint caused by the integration of a multi-energy system into the power grid. max P represents the maximum acceptable photovoltaic capacity of the power grid. h,i Let P represent the output of the i-th adjustable hydroelectric power station. h,i,min P h,i,max Let P represent the minimum and maximum output of the i-th adjustable hydropower station, respectively. When the adjustable hydropower station is a horizontal pumped storage power station, P P,max N represents the maximum output of the photovoltaic power station in a multi-energy system. pv N represents the installed capacity of a multi-energy photovoltaic power station. pv,max P represents the maximum allowable installed capacity of a multi-energy photovoltaic power station. f(t) Q represents the load of the multi-energy system at time t. Pt This represents the amount of wind and solar power curtailment in the multi-energy system at time t.

6. An optimized scheduling device for a water-solar hybrid system combining optical power prediction, characterized in that, include: The photovoltaic output data acquisition module is used to collect the operating characteristic data, meteorological characteristic data and time characteristic data of the photovoltaic system, and to preprocess the data, including missing value reconstruction, smoothing of non-stationary components, standardization or normalization, and conversion of categorical data into numerical codes to construct a multidimensional time series dataset for training. The prediction module is used to construct a light power prediction model based on a long short-term memory model. The model is trained using the multi-dimensional time series dataset, and the nonlinear relationship and time-dependent features between light power and meteorological factors are extracted. The model parameters are dynamically adjusted through an adaptive moment estimation optimization algorithm, and the root mean square error and mean absolute error are used as loss functions for model optimization. The hydropower station data acquisition module is used to collect real-time data on water level, flow rate, and power generation of the hydropower station, as well as the light intensity, temperature, relative humidity, and power generation of the photovoltaic power station, and transmit the data to the dispatch center. The execution scheduling optimization module is used to construct a scheduling optimization model for the hydro-solar hybrid system. It uses a non-dominated sorting genetic algorithm to configure the photovoltaic power generation capacity, optimizes the scheduling so that the photovoltaic installed capacity maximizes power generation and minimizes curtailment under the constraints. It also adjusts the hydropower output according to the system operating status to achieve real-time scheduling and control of hydropower and photovoltaic power.

7. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-5.

9. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method of any one of claims 1-5.

Images