Intelligent scheduling method and system for intelligent mining of water-solar complementary characteristics

Abstract

This invention discloses an intelligent scheduling method and system for intelligently mining the complementary characteristics of solar and hydropower, belonging to the field of renewable energy scheduling technology. Specifically, it includes: acquiring operational data and meteorological data from distributed photovoltaic power stations and cascaded small hydropower stations, and performing preprocessing; analyzing the dynamic complementary relationship between distributed photovoltaic and cascaded small hydropower stations, and constructing a dynamic solar-hydropower complementary characteristic model; based on the dynamic solar-hydropower complementary characteristic model, and combining photovoltaic output fluctuations, the regulation capacity of cascaded small hydropower stations, and reservoir scheduling constraints, constructing a joint power generation model for distributed photovoltaic and cascaded small hydropower stations, and solving for the optimal scheduling scheme through an intelligent optimization scheduling algorithm; establishing a joint operation control system, dynamically adjusting the output plan according to the optimal scheduling scheme, and optimizing the intelligent optimization scheduling algorithm and the joint power generation model through a feedback mechanism; achieving intelligent mining and optimized scheduling of solar-hydropower complementary characteristics, improving energy utilization efficiency and system stability.

Description

Technical Field

[0001] This invention belongs to the field of renewable energy dispatching technology, specifically an intelligent dispatching method and system for intelligently exploiting the complementary characteristics of water and solar power. Background Technology

[0002] With the rapid development of distributed photovoltaic (PV) power and cascaded small hydropower, the joint operation of hydropower and PV power plants has become an important way to solve the problems of energy volatility and intermittency. Distributed PV is greatly affected by weather, resulting in fluctuating and intermittent power output, while cascaded small hydropower has the characteristics of strong regulation capabilities and high energy storage efficiency. How to achieve joint operation control and intelligent scheduling of the two has become an important issue in the field of renewable energy. In existing technologies, hydropower-PV complementary scheduling mostly adopts fixed rules or simple optimization models, failing to fully explore the dynamic complementary characteristics of distributed PV and cascaded small hydropower, resulting in unsatisfactory scheduling effects and low energy utilization.

[0003] For example, Chinese patent CN117543721B discloses an optimized scheduling method, device, equipment, and medium for a cascade hydropower-wind-solar hybrid system. The method includes: establishing constraints for scheduling the cascade hydropower-wind-solar hybrid system, including constraints on the power characteristics, hydraulic characteristics, and hydropower coupling characteristics of the cascade hydropower stations; constraints on the hydropower-wind-solar coupling characteristics and output characteristics of the cascade hydropower-wind-solar hybrid system; constructing a scheduling model for the cascade hydropower-wind-solar hybrid system based on the constraints, with the objective function being the maximization of the sum of the medium- and long-term scheduling utility and the short-term scheduling utility of the cascade hydropower-wind-solar hybrid system; and solving the scheduling model using a linearization method to obtain the optimized scheduling result. This technical solution can provide a power generation scheme for a cascade hydropower-wind power-solar hybrid system that satisfies the constraints, improve the overall system utility, reduce wind and solar curtailment, realize resource integration of new energy power generation systems, and ensure the safe and stable operation of the power grid.

[0004] The existing technologies described above have the following problems: they rely on static constraints and objective functions, lacking adaptability to dynamic changes in the system; they employ traditional optimization methods, such as linearization methods, and do not fully utilize data-driven approaches; they do not introduce real-time feedback mechanisms and online optimization techniques, making it impossible to dynamically adjust scheduling strategies based on actual operational data; although they consider the constraints of the hydro-wind-solar hybrid system, they do not deeply explore the dynamic complementary characteristics between water, wind, and solar; they mainly focus on medium- and long-term scheduling utility and short-term scheduling utility, without fully considering scheduling needs at multiple time scales, such as ultra-short-term, short-term, medium-term, and long-term. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention proposes an intelligent scheduling method and system for intelligently mining the complementary characteristics of solar and hydropower. The method acquires and preprocesses operational and meteorological data from distributed photovoltaic power plants and cascaded small hydropower stations. It analyzes the dynamic complementary relationship between distributed photovoltaic and cascaded small hydropower stations, constructing a dynamic solar-hydropower complementary characteristic model. Based on this model, and considering photovoltaic output fluctuations, the regulation capacity of cascaded small hydropower stations, and reservoir scheduling constraints, a joint power generation model for distributed photovoltaic and cascaded small hydropower stations is constructed. The optimal scheduling scheme is then solved using an intelligent optimization scheduling algorithm. A joint operation control system is established, dynamically adjusting the output plan according to the optimal scheduling scheme. The intelligent optimization scheduling algorithm and joint power generation model are optimized through a feedback mechanism. This achieves intelligent mining and optimized scheduling of the complementary characteristics of solar and hydropower, improving energy utilization efficiency and system stability.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] Intelligent scheduling methods for intelligently mining the characteristics of water-solar hybridity include:

[0008] Step S1: Obtain operational data from distributed photovoltaic power stations and cascade small hydropower stations. Simultaneously, acquire meteorological data and perform preprocessing to obtain operational characteristic data.

[0009] Step S2: Based on the operational feature data, use machine learning algorithms to analyze the output characteristics of distributed photovoltaic and cascaded small hydropower, explore the dynamic complementary relationship between distributed photovoltaic and cascaded small hydropower, and construct a dynamic hydro-photovoltaic complementary characteristic model based on the dynamic complementary relationship and combined with deep reinforcement learning algorithms to predict the photovoltaic output and cascaded small hydropower regulation capacity at different time scales.

[0010] Step S3: Based on the water-solar complementary characteristic model, construct a joint power generation model of distributed photovoltaic and cascade small hydropower, and use the intelligent optimization scheduling algorithm to solve the optimal scheduling scheme of the joint power generation model, and output the optimal scheduling scheme to the joint operation control system;

[0011] Step S4: Based on the optimal scheduling scheme, the joint operation control system dynamically adjusts the output plans of distributed photovoltaic and cascade small hydropower, monitors the operation effect in real time, collects actual operation data, and compares the actual operation data with the prediction results through a feedback mechanism. Based on the comparison results, online learning technology is used to dynamically optimize the intelligent optimization scheduling algorithm and the joint power generation model.

[0012] Specifically, step S2 includes the following steps:

[0013] S2.1: Obtain runtime characteristic data The running feature data is input into a pre-loaded random forest for training to obtain a trained decision tree model. Denotes the i-th running characteristic, and N represents the number of running features;

[0014] S2.2: Based on each feature In the trained decision tree model, the importance score of each feature is calculated by combining the split nodes and corresponding split gains, the interactions between features, and the time decay. At the same time, for Sort the data and select the top M features to obtain the filtered running feature data. ,and ;

[0015] S2.3: Load the pre-built random forest model, input the filtered operational feature data into the pre-built random forest model for training, and obtain the prediction results of photovoltaic power output and cascade small hydropower regulation capacity.

[0016] Specifically, step S2 further includes the following steps:

[0017] S2.4: Use clustering algorithms to perform cluster analysis on the prediction results of photovoltaic power output and the regulation capacity of cascade small hydropower, and identify the complementary patterns between photovoltaic power output and the regulation capacity of cascade small hydropower.

[0018] S2.5: Based on the identification results, extract the complementary relationship between photovoltaic and small hydropower under different meteorological conditions, and combine reinforcement learning algorithms to simulate the changes in the complementary relationship under different meteorological conditions and scheduling strategies;

[0019] S2.6: Combining the prediction results of the random forest model and the simulation results of reinforcement learning, a dynamic water-solar complementary characteristic model is constructed, and the dynamic water-solar complementary characteristic model is used to predict the photovoltaic output and the regulation capacity of cascade small hydropower at different time scales.

[0020] S2.7: Collect real-time operating data and input the real-time operating data into the dynamic water-solar complementary characteristic model. Use an online learning algorithm to update the parameters of the dynamic water-solar complementary characteristic model and obtain the updated dynamic water-solar complementary characteristic model.

[0021] Specifically, the steps in S2.4 include:

[0022] S2.41: Align the prediction results of photovoltaic power output and the regulation capacity of cascade small hydropower by time point, organize them into a prediction result dataset, and standardize the data in the prediction result dataset. Each row in the prediction result dataset represents the predicted value at a time point.

[0023] S2.42: Set the number of clusters, and use a clustering algorithm to cluster the data in the standardized prediction result dataset, and output the cluster label for each data point;

[0024] S2.43: For each cluster, the mean of the data points in each cluster is taken as the cluster center point, and the photovoltaic output and cascade hydropower regulation capacity values ​​of the cluster center point are extracted.

[0025] If the center point of the cluster shows high photovoltaic output and low small hydropower regulation capacity, it means that under the current circumstances, photovoltaic power is the source of power generation and small hydropower is a supplementary power source.

[0026] If the center point of the cluster shows low photovoltaic output and high small hydropower regulation capacity, it means that under the current conditions, small hydropower is the source of power generation and photovoltaic power is a supplementary power source.

[0027] S2.44: Based on the judgment result of S2.43, define the types of complementary modes, including high photovoltaic + low small hydropower and low photovoltaic + high small hydropower;

[0028] S2.45: Based on the clustering results and complementary patterns, draw a scatter plot of photovoltaic power output and the regulation capacity of cascade small hydropower, and use different colors to represent different clusters. At the same time, mark the center point of each cluster in the scatter plot.

[0029] Specifically, the steps of S2.5 include:

[0030] S2.51: Associate the center point of each cluster with the corresponding meteorological conditions to extract the complementary relationship between photovoltaic and small hydropower under different meteorological conditions;

[0031] S2.52: Based on the complementary relationship between photovoltaic and small hydropower under different meteorological conditions, a reinforcement learning environment is constructed, including defining a state space, an action space, and a reward function; the state space includes the current meteorological conditions, photovoltaic output, small hydropower regulation capacity, and reservoir water level; the action space includes reservoir water release strategy and photovoltaic output adjustment strategy.

[0032] S2.53: Load the pre-trained reinforcement learning model and input different meteorological conditions and scheduling strategies into the pre-trained reinforcement learning model to simulate changes in complementary relationships; the scheduling strategies include reservoir water release strategies and photovoltaic power output adjustment strategies.

[0033] S2.54: Based on the simulation results, extract the variation law of the complementary relationship between photovoltaic and small hydropower under different meteorological conditions and dispatch strategies.

[0034] Specifically, the steps in S2.6 include:

[0035] S2.61: Align the prediction results of the random forest model with the simulation results of reinforcement learning by timestamp and perform feature fusion to obtain a high-dimensional feature dataset;

[0036] S2.62: Load the pre-built machine learning model and train the pre-built machine learning model using a high-dimensional feature dataset to obtain a dynamic water-photovoltaic complementary characteristic model;

[0037] S2.63: Obtain meteorological conditions and reservoir water level data at different time scales, and input them into the dynamic water-solar complementary characteristic model to obtain photovoltaic power output and cascade small hydropower regulation capacity at different time scales.

[0038] Specifically, step S3 includes the following steps:

[0039] S3.1: Collect photovoltaic power output fluctuation data and cascade small hydropower regulation capacity data;

[0040] S3.2: Set reservoir scheduling constraints ,in, Indicates the reservoir water level. and These represent the lower and upper limits of the reservoir water level, respectively. Indicates the reservoir flow rate. and These represent the lower and upper limits of the reservoir's flow rate, respectively.

[0041] S3.3: Setting the multi-objective optimization function ,in, This indicates power generation efficiency, and Cost indicates the cost of generating electricity. Indicates carbon emissions, , , Indicates the weighting coefficient;

[0042] S3.4: Based on the characteristic model of water-solar complementarity, combined with photovoltaic power output fluctuation data, cascade small hydropower regulation capacity data and reservoir scheduling constraints, a joint power generation model is constructed;

[0043] S3.5: Use intelligent optimization scheduling algorithm to solve the joint generation model and obtain the optimal scheduling scheme;

[0044] S3.6: Transmit the optimal scheduling scheme to the joint operation control system.

[0045] Specifically, the steps in S3.5 include:

[0046] S3.51: Set the parameters of the differential evolution algorithm, including population size, scaling factor, crossover probability and maximum number of iterations;

[0047] S3.52: Randomly generate an initial population, wherein each individual in the population... This represents a scheduling scheme;

[0048] S3.53: Based on the multi-objective optimization function in S3.3, calculate the objective function value for each individual. ;

[0049] S3.54: Pass Generate mutated individuals The mutant individuals and the original individuals were cross-operated to generate experimental individuals. ,in, , , This represents three randomly selected distinct individuals. Indicates the scaling factor;

[0050] S3.55: Comparative test individuals and the original individual The objective function value;

[0051] like Then select test individuals Moving into the next generation;

[0052] like Then select the original individual. Moving into the next generation;

[0053] S3.56: If the current iteration count reaches the maximum iteration count, terminate the differential evolution algorithm and select the individual with the highest objective function value in the population as the optimal scheduling scheme.

[0054] Specifically, the joint operation control system includes grid-connected joint control, fast and smooth control, off-grid coordinated control, and advanced predictive control functions.

[0055] The intelligent scheduling system for intelligent mining of the characteristics of hydro-solar complementarity includes: a data acquisition module, a complementary analysis module, a joint power generation model construction module, and a feedback optimization module;

[0056] The data acquisition module is used to acquire multi-source operation data from distributed photovoltaic power stations and cascade small hydropower systems, perform intelligent preprocessing, and store the preprocessed data in a distributed database.

[0057] The complementary analysis module uses machine learning algorithms and reinforcement learning techniques to analyze the output characteristics of photovoltaic and small hydropower, explore dynamic complementary relationships, and construct a dynamic hydro-photovoltaic complementary characteristic model.

[0058] The joint power generation model construction module constructs a joint power generation model based on the water-solar complementary characteristic model, and designs a multi-objective optimization function to solve the optimal scheduling scheme.

[0059] The feedback optimization module dynamically adjusts the output plan according to the optimal scheduling scheme, monitors the operation effect in real time, and optimizes the model and algorithm through the feedback mechanism.

[0060] Compared with the prior art, the beneficial effects of the present invention are:

[0061] 1. This invention proposes an intelligent scheduling system for intelligent mining of the characteristics of water-solar complementarity, and optimizes and improves the architecture, operation steps and processes. The system has the advantages of simple process, low investment and operation costs and low production and working costs.

[0062] 2. This invention proposes an intelligent scheduling method for intelligently mining the characteristics of water-solar complementarity. This intelligent scheduling method integrates distributed photovoltaic power stations, cascaded small hydropower stations, and meteorological data, and uses machine learning and deep reinforcement learning algorithms to mine the characteristics of water-solar complementarity. It achieves accurate prediction of photovoltaic output and hydropower regulation capacity at different time scales, thereby improving the prediction accuracy and flexibility of the energy system.

[0063] 3. This invention proposes an intelligent scheduling method for intelligent mining of the characteristics of hydro-solar complementarity. By constructing a joint power generation model and applying an intelligent optimization scheduling algorithm, the optimal hydro-solar complementary power generation scheduling scheme can be formulated, and the operation plan can be adjusted in real time to adapt to actual changes. At the same time, online learning technology is used to continuously optimize the model and algorithm, ensuring the efficient and stable operation of the energy system and improving energy utilization efficiency. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of the intelligent scheduling method for intelligent mining of the water-solar complementary characteristics of the present invention;

[0065] Figure 2 This is a flowchart illustrating the principle of the intelligent scheduling method for intelligent mining of the water-solar complementary characteristics of the present invention.

[0066] Figure 3 This is a diagram of the intelligent scheduling system architecture for intelligent mining of the water-solar complementary characteristics of the present invention. Detailed Implementation

[0067] Example 1

[0068] Please see Figure 1 and Figure 2 The present invention provides an embodiment of an intelligent scheduling method for intelligent mining of water-solar complementary characteristics, comprising the following steps:

[0069] Step S1: Obtain operational data from distributed photovoltaic power stations and cascade small hydropower stations. Simultaneously, acquire meteorological data and perform preprocessing to obtain operational characteristic data.

[0070] Furthermore, the specific steps of step S1 include:

[0071] (1) Collect real-time operation data from photovoltaic power stations and small hydropower systems, and obtain meteorological data from meteorological stations or meteorological forecasting systems, and store the collected operation data and meteorological data in a distributed database; the meteorological data includes light intensity, temperature, rainfall, etc.

[0072] (2) Clean the collected real-time operational data and meteorological data, and process missing and outlier values ​​to ensure data quality;

[0073] (3) Integrate meteorological data, geographic information system data and real-time operation data to generate a high-dimensional feature dataset, and normalize the high-dimensional feature dataset to obtain operation feature data.

[0074] Step S2: Based on the operational feature data, use machine learning algorithms to analyze the output characteristics of distributed photovoltaic and cascaded small hydropower, explore the dynamic complementary relationship between distributed photovoltaic and cascaded small hydropower, and construct a dynamic hydro-photovoltaic complementary characteristic model based on the dynamic complementary relationship and combined with deep reinforcement learning algorithms to predict the photovoltaic output and cascaded small hydropower regulation capacity at different time scales.

[0075] Step S3: Based on the characteristic model of water-solar complementarity, combined with the photovoltaic power output fluctuation, the regulation capacity of cascade small hydropower and the reservoir scheduling constraints, a joint power generation model of distributed photovoltaic and cascade small hydropower is constructed. The optimal scheduling scheme of the joint power generation model is solved by intelligent optimization scheduling algorithm, and the optimal scheduling scheme is output to the joint operation control system.

[0076] The joint operation control system includes grid-connected joint control, fast and smooth control, off-grid coordinated control, and advanced predictive control functions;

[0077] Step S4: Based on the optimal scheduling scheme, the joint operation control system dynamically adjusts the output plans of distributed photovoltaic and cascade small hydropower, monitors the operation effect in real time, collects actual operation data, and compares the actual operation data with the prediction results through a feedback mechanism. Based on the comparison results, online learning technology is used to dynamically optimize the intelligent optimization scheduling algorithm and the joint power generation model.

[0078] Furthermore, the specific steps of step S4 include:

[0079] (1) Analyze the optimal scheduling scheme, obtain the output plan of distributed photovoltaic and cascade small hydropower, dynamically adjust the output plan of distributed photovoltaic and cascade small hydropower according to the current meteorological conditions, load demand and equipment status, and output the adjusted output plan to the distributed photovoltaic power station and cascade small hydropower system to execute the scheduling;

[0080] (2) Collect actual power output data, meteorological data, and reservoir water level data of distributed photovoltaic and cascade small hydropower, and store the collected actual operation data in a database or data warehouse;

[0081] (3) Compare the actual operating data with the prediction results, and obtain the error by calculating the mean of the differences between all actual operating data and prediction results;

[0082] (4) The parameters of the intelligent optimization scheduling algorithm and the joint generation model are updated using the stochastic gradient descent algorithm, and the intelligent optimization scheduling algorithm and the joint generation model are optimized according to the updated parameters to achieve rolling optimization. The stochastic gradient descent algorithm is the prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0083] The specific steps of step S2 include:

[0084] S2.1: Obtain runtime characteristic data The running feature data is input into a pre-loaded random forest for training to obtain a trained decision tree model. Denotes the i-th running characteristic, and N represents the number of running features;

[0085] S2.2: Based on each feature In the trained decision tree model, the importance score of each feature is calculated by combining the split nodes and corresponding split gains, the interactions between features, and the time decay. At the same time, for Sort the data and select the top M features to obtain the filtered running feature data. ,and ,in, The score represents the importance of feature i, and n represents the number of decision trees. This represents the split node of the j-th decision tree. This represents the splitting gain of feature i at node t. This represents the weight coefficient of feature i at node t. This represents the interaction coefficient between feature i and feature k. This represents the time decay factor; the operational characteristic data includes photovoltaic power generation, light intensity, reservoir water level, and flow rate.

[0086] Furthermore, Let represent the weight coefficient of feature i at node t, and satisfy . ,in, Indicates the depth of node t. This represents the number of samples at node t.

[0087] Furthermore, Let represent the interaction coefficient between feature i and feature k, and satisfy . ,in, Represents the interaction strength coefficient. This represents the correlation coefficient between feature i and feature k.

[0088] Furthermore, Let represent the time decay factor, and satisfy . ,in, Indicates the decay rate. Indicates the current time. This represents the timestamp of node t.

[0089] It should be noted that the split gain measures the splitting effect of feature i at the current node t. The feature importance is dynamically adjusted by splitting depth and number of samples to avoid interference from excessively deep nodes or too few samples on the feature importance assessment. The synergistic effect between features is introduced to enhance the ability to assess feature interactions. The time decay factor can dynamically adjust feature importance to make the model more adaptable to the changing trends of the data.

[0090] S2.3: Load the pre-built random forest model, input the filtered operational feature data into the pre-built random forest model for training, and obtain the prediction results of photovoltaic power output and cascade small hydropower regulation capacity;

[0091] Furthermore, the specific steps in S2.3 include:

[0092] (1) Load the pre-built random forest model framework from the model storage path and load the model using the machine learning library Scikit-learn;

[0093] (2) The selected operational feature data are divided into feature matrix and target variables. The feature matrix includes light intensity, temperature, reservoir water level and flow rate, and the target variables include photovoltaic output and cascade small hydropower regulation capacity.

[0094] (3) Use the train_test_split function of Scikit-learn to divide the filtered running feature data into training and test sets in an 8:2 ratio;

[0095] (4) Input the training set data into the random forest model for training to obtain the trained random forest model;

[0096] (5) Use the test set data to make predictions and obtain the prediction results of photovoltaic power output and the regulation capacity of cascade small hydropower. Store the prediction results. For the random forest model, the prediction result is the average of the prediction results of all decision trees. The random forest model is the prior art in this field and is not an inventive solution of this application. It will not be described in detail here.

[0097] S2.4: Use clustering algorithms to perform cluster analysis on the prediction results of photovoltaic power output and the regulation capacity of cascade small hydropower, and identify the complementary patterns between photovoltaic power output and the regulation capacity of cascade small hydropower.

[0098] S2.5: Based on the identification results, extract the complementary relationship between photovoltaic and small hydropower under different meteorological conditions, and combine reinforcement learning algorithms to simulate the changes in the complementary relationship under different meteorological conditions and scheduling strategies;

[0099] S2.6: Combining the prediction results of the random forest model and the simulation results of reinforcement learning, a dynamic water-solar complementary characteristic model is constructed, and the dynamic water-solar complementary characteristic model is used to predict the photovoltaic output and the regulation capacity of cascade small hydropower at different time scales.

[0100] S2.7: Collect real-time operating data and input the real-time operating data into the dynamic water-solar complementary characteristic model. Use an online learning algorithm to update the parameters of the dynamic water-solar complementary characteristic model and obtain the updated dynamic water-solar complementary characteristic model.

[0101] Furthermore, the specific steps in S2.7 include:

[0102] (1) Real-time collection of operation data of distributed photovoltaic and cascade small hydropower, and input of real-time operation data into dynamic water-photovoltaic complementary characteristic model for model updating and prediction;

[0103] (2) The parameters of the dynamic water-photovoltaic complementary characteristic model are updated using the stochastic gradient descent algorithm, and the performance of the updated dynamic water-photovoltaic complementary characteristic model is evaluated using real-time running data. The mean square error and mean absolute error are calculated. The formulas for calculating the mean square error and mean absolute error are existing technologies in this field and are not the inventive solutions of this application. They will not be elaborated here.

[0104] (3) Store the updated dynamic water-solar complementary characteristic model in the model library. At the same time, deploy the updated model to the joint operation control system for real-time prediction and scheduling.

[0105] The specific steps in S2.4 include:

[0106] S2.41: Align the prediction results of photovoltaic power output and the regulation capacity of cascade small hydropower by time point, organize them into a prediction result dataset, and standardize the data in the prediction result dataset. Each row in the prediction result dataset represents the predicted value at a time point.

[0107] S2.42: Set the number of clusters, and use a clustering algorithm to cluster the data in the standardized prediction result dataset, and output the cluster label of each data point. The clustering algorithm used is the k-means algorithm, but the k-means algorithm is the prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0108] S2.43: For each cluster, the mean of the data points in each cluster is taken as the cluster center point, and the photovoltaic output and cascade hydropower regulation capacity values ​​of the cluster center point are extracted.

[0109] If the center point of the cluster shows high photovoltaic output and low small hydropower regulation capacity, it means that under the current circumstances, photovoltaic power is the source of power generation and small hydropower is a supplementary power source.

[0110] If the center point of the cluster shows low photovoltaic output and high small hydropower regulation capacity, it means that under the current conditions, small hydropower is the source of power generation and photovoltaic power is a supplementary power source.

[0111] In this invention, if the output power of a distributed photovoltaic power station is greater than or equal to 70% of its rated power, it is considered to have high photovoltaic output. For example, if a photovoltaic power station with a rated power of 1000 kW has an actual output power of 700 kW or more, it is considered to be in a high photovoltaic output state. If the adjustable power generation range of a small hydropower station is less than 30% of its rated power and the adjustment response time exceeds 30 minutes, it is considered to have low small hydropower adjustment capability. For example, if a small hydropower station with a rated power of 500 kW has an adjustable power range of less than 150 kW and a response time of 1 hour, it is considered to have low small hydropower adjustment capability.

[0112] S2.44: Based on the judgment result of S2.43, define the types of complementary modes, including high photovoltaic + low small hydropower and low photovoltaic + high small hydropower;

[0113] S2.45: Based on the clustering results and complementary patterns, draw a scatter plot of photovoltaic power output and the regulation capacity of cascade small hydropower, and use different colors to represent different clusters. At the same time, mark the center point of each cluster in the scatter plot.

[0114] The specific steps in S2.5 include:

[0115] S2.51: Associate the center point of each cluster with the corresponding meteorological conditions to extract the complementary relationship between photovoltaic and small hydropower under different meteorological conditions;

[0116] Furthermore, the specific steps of S2.51 include:

[0117] (1) Extract the center point data of each cluster from the clustering results, including photovoltaic power output and cascade hydropower regulation capacity;

[0118] (2) Obtain meteorological condition data corresponding to the timestamp of the cluster center point, including light intensity, temperature, and rainfall;

[0119] (3) Align cluster center point data with meteorological condition data by timestamp to ensure data consistency;

[0120] (4) Combine the cluster center point data and meteorological condition data into a single dataset, with each data point including cluster center point features and corresponding meteorological conditions;

[0121] (5) The correlation coefficient analysis method is used to analyze the relationship between the cluster center point characteristics and meteorological conditions, and the correlation analysis results are stored. The correlation coefficient analysis method is the prior art in this field and is not an inventive solution of this application. It will not be described in detail here.

[0122] S2.52: Based on the complementary relationship between photovoltaic and small hydropower under different meteorological conditions, a reinforcement learning environment is constructed, including defining a state space, an action space, and a reward function; the state space includes the current meteorological conditions, photovoltaic output, small hydropower regulation capacity, and reservoir water level; the action space includes reservoir water release strategy and photovoltaic output adjustment strategy.

[0123] S2.53: Load the pre-trained reinforcement learning model and input different meteorological conditions and scheduling strategies into the pre-trained reinforcement learning model to simulate the changes in complementary relationships; the scheduling strategies include reservoir water release strategies and photovoltaic power output adjustment strategies. The reinforcement learning model is existing technology in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0124] Among these strategies, reservoir water release and photovoltaic (PV) output adjustment are key tactics for optimizing scheduling and resource allocation in a cascade hydropower-wind-solar hybrid system. Reservoir water release involves regulating the amount of water released from the reservoir to control the power generation capacity of the cascade hydropower stations, thereby meeting electricity demand, optimizing water resource utilization, and maximizing overall system benefits. PV output adjustment involves adjusting the output of the PV power stations to complement the output of the cascade hydropower stations, thus optimizing the overall system output characteristics. Because PV output is significantly affected by weather conditions, exhibiting fluctuations and intermittency, while hydropower station output can be flexibly adjusted through reservoir water release, the synergistic effect of these two strategies allows for hydropower-solar complementarity, optimizing the overall system output characteristics. For example, when PV output is high, the output of the hydropower stations can be reduced to conserve water resources; conversely, when PV output is low, the output of the hydropower stations can be increased to compensate for insufficient PV power generation.

[0125] S2.54: Based on the simulation results, extract the variation law of the complementary relationship between photovoltaic and small hydropower under different meteorological conditions and dispatch strategies.

[0126] The specific steps in S2.6 include:

[0127] S2.61: Align the prediction results of the random forest model with the simulation results of reinforcement learning by timestamp and perform feature fusion to obtain a high-dimensional feature dataset. For example, combine features such as photovoltaic output, cascade hydropower regulation capacity, meteorological conditions, and reservoir water level into one dataset.

[0128] S2.62: Load the pre-built machine learning model and train the pre-built machine learning model using a high-dimensional feature dataset to obtain a dynamic water-photovoltaic complementary characteristic model. The machine learning model is existing technology in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0129] S2.63: Obtain meteorological conditions and reservoir water level data at different time scales, and input them into the dynamic water-solar complementary characteristic model to obtain photovoltaic power output and cascade small hydropower regulation capacity at different time scales.

[0130] The specific steps of step S3 include:

[0131] S3.1: Collect photovoltaic power output fluctuation data and cascade small hydropower regulation capacity data;

[0132] S3.2: Set reservoir scheduling constraints ,in, Indicates the reservoir water level. and These represent the lower and upper limits of the reservoir water level, respectively. Indicates the reservoir flow rate. and These represent the lower and upper limits of the reservoir's flow rate, respectively.

[0133] S3.3: Setting the multi-objective optimization function ,in, This indicates power generation efficiency, and Cost indicates the cost of generating electricity. Indicates carbon emissions, , , Indicates the weighting coefficient;

[0134] S3.4: Based on the characteristic model of water-solar complementarity, combined with photovoltaic power output fluctuation data, cascade small hydropower regulation capacity data and reservoir scheduling constraints, a joint power generation model is constructed;

[0135] The process of building a combined power generation model includes:

[0136] (1) Linear programming is selected as the joint generation model;

[0137] (2) Set model parameters, including objective function, constraints, and decision variables;

[0138] (3) Obtain the joint power generation model.

[0139] S3.5: Use intelligent optimization scheduling algorithm to solve the joint generation model and obtain the optimal scheduling scheme;

[0140] S3.6: Transmit the optimal scheduling scheme to the joint operation control system.

[0141] The specific steps in S3.5 include:

[0142] S3.51: Set the parameters of the differential evolution algorithm, including population size, scaling factor, crossover probability and maximum number of iterations;

[0143] S3.52: Randomly generate an initial population, wherein each individual in the population... This represents a scheduling scheme;

[0144] S3.53: Based on the multi-objective optimization function in S3.3, calculate the objective function value for each individual. ;

[0145] S3.54: Pass Generate mutated individuals The mutant individuals and the original individuals were cross-operated to generate experimental individuals. ,in, , , This represents three randomly selected distinct individuals. Indicates the scaling factor;

[0146] S3.55: Comparative test individuals and the original individual The objective function value;

[0147] like Then select test individuals Moving into the next generation;

[0148] like Then select the original individual. Moving into the next generation;

[0149] S3.56: If the current iteration count reaches the maximum iteration count, terminate the differential evolution algorithm and select the individual with the highest objective function value in the population as the optimal scheduling scheme.

[0150] Example 2

[0151] Please see Figure 3 Another embodiment of the present invention provides an intelligent scheduling system for intelligent mining of water-solar complementary characteristics, comprising:

[0152] Data acquisition module, complementary analysis module, joint power generation model construction module, feedback optimization module;

[0153] The data acquisition module is used to acquire multi-source operation data from distributed photovoltaic power stations and cascade small hydropower systems, perform intelligent preprocessing, and store the preprocessed data in a distributed database.

[0154] The complementary analysis module uses machine learning algorithms and reinforcement learning techniques to analyze the output characteristics of photovoltaic and small hydropower, explore dynamic complementary relationships, and construct a dynamic hydro-photovoltaic complementary characteristic model.

[0155] The module for constructing a combined power generation model is based on the characteristic model of water-solar complementarity. It constructs a combined power generation model and designs a multi-objective optimization function to solve for the optimal scheduling scheme.

[0156] The feedback optimization module dynamically adjusts the output plan based on the optimal scheduling scheme, monitors the operational effect in real time, and optimizes the model and algorithm through the feedback mechanism.

[0157] The complementary analysis module includes: a feature extraction unit, a machine learning analysis unit, a reinforcement learning modeling unit, and a prediction unit;

[0158] The feature extraction unit is used to extract key features from preprocessed data, such as the light dependence of photovoltaic power output and the water flow regulation characteristics of cascade small hydropower.

[0159] The machine learning analysis unit uses machine learning algorithms to analyze the output characteristics of photovoltaic and small hydropower.

[0160] The reinforcement learning modeling unit uses reinforcement learning algorithms to simulate the changes in complementary relationships under different meteorological conditions and scheduling strategies, and constructs a dynamic water-solar complementary characteristic model.

[0161] The prediction unit predicts photovoltaic output and cascade hydropower regulation capacity at different time scales based on a dynamic hydro-solar complementary characteristic model. These different time scales include short-term, medium-term, and long-term.

[0162] The joint power generation model construction module includes: a model construction unit, an optimization function design unit, an intelligent optimization scheduling unit, and a scheme output unit;

[0163] The model building unit combines the fluctuations in photovoltaic output, the regulation capacity of cascade small hydropower stations, and reservoir scheduling constraints to construct a joint power generation model;

[0164] The optimization function design unit designs multi-objective optimization functions, where the objectives include maximizing power generation efficiency, minimizing power generation cost, and reducing carbon emissions.

[0165] The intelligent optimization scheduling unit uses intelligent optimization scheduling algorithms to solve for the optimal scheduling scheme.

[0166] The scheme output unit is used to output the optimal scheduling scheme to the joint operation control system.

[0167] The feedback optimization module includes: a runtime control unit, a real-time monitoring unit, a feedback comparison unit, and a model optimization unit;

[0168] The operation control unit dynamically adjusts the output plans of distributed photovoltaic and cascade small hydropower according to the optimal scheduling scheme;

[0169] The real-time monitoring unit is used to collect system operation data in real time, such as actual power generation, equipment status, and weather changes;

[0170] The feedback comparison unit is used to compare the actual operating data with the prediction results, calculate the prediction error, and analyze the source of the error.

[0171] The model optimization unit utilizes online learning technology to dynamically update the hydro-solar complementary characteristic model and the joint power generation model, and optimizes the intelligent scheduling algorithm.

[0172] In summary, through the design of the data acquisition module, complementary analysis module, joint power generation model construction module, feedback optimization module, and their corresponding units, the system achieves a complete closed loop from data acquisition, model construction, optimized scheduling to feedback optimization, which is logically sound and highly efficient.

[0173] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments under the guidance of the present invention without departing from the spirit and scope of the present invention. All of these variations are within the protection scope of the present invention.

Claims

1. An intelligent scheduling method for intelligently mining the characteristics of water-solar hybridity, characterized in that, include: Step S1: Obtain operational data from distributed photovoltaic power stations and cascade small hydropower stations. Simultaneously, acquire meteorological data and perform preprocessing to obtain operational characteristic data. Step S2: Based on the operational feature data, use machine learning algorithms to analyze the output characteristics of distributed photovoltaic and cascaded small hydropower, explore the dynamic complementary relationship between distributed photovoltaic and cascaded small hydropower, and construct a dynamic hydro-photovoltaic complementary characteristic model based on the dynamic complementary relationship and combined with deep reinforcement learning algorithms to predict the photovoltaic output and cascaded small hydropower regulation capacity at different time scales. Step S3: Based on the water-solar complementary characteristic model, construct a joint power generation model of distributed photovoltaic and cascade small hydropower, and use the intelligent optimization scheduling algorithm to solve the optimal scheduling scheme of the joint power generation model, and output the optimal scheduling scheme to the joint operation control system; Step S4: According to the optimal scheduling scheme, the joint operation control system dynamically adjusts the output plan of distributed photovoltaic and cascade small hydropower, monitors the operation effect in real time, collects actual operation data, and compares the actual operation data with the prediction results through the feedback mechanism. Based on the comparison results, the intelligent optimization scheduling algorithm and joint power generation model are dynamically optimized using online learning technology. The specific steps of step S2 include: S2.1: Obtain runtime characteristic data The running feature data is input into a pre-loaded random forest for training to obtain a trained decision tree model. Indicates the first i Each operating characteristic, and N represents the number of running features; S2.2: Based on each feature In the trained decision tree model, the importance score of each feature is calculated by combining the split nodes and corresponding split gains, the interactions between features, and the time decay. At the same time, for Sort the data and select the top M features to obtain the filtered running feature data. ,and ; S2.3: Load the pre-built random forest model, input the filtered operational feature data into the pre-built random forest model for training, and obtain the prediction results of photovoltaic power output and cascade small hydropower regulation capacity; The specific steps of step S2 also include: S2.4: Use clustering algorithms to perform cluster analysis on the prediction results of photovoltaic power output and the regulation capacity of cascade small hydropower, and identify the complementary patterns between photovoltaic power output and the regulation capacity of cascade small hydropower. S2.5: Based on the identification results, extract the complementary relationship between photovoltaic and small hydropower under different meteorological conditions, and combine reinforcement learning algorithms to simulate the changes in the complementary relationship under different meteorological conditions and scheduling strategies; S2.6: Combining the prediction results of the random forest model and the simulation results of reinforcement learning, a dynamic water-solar complementary characteristic model is constructed, and the dynamic water-solar complementary characteristic model is used to predict the photovoltaic output and the regulation capacity of cascade small hydropower at different time scales. S2.7: Collect real-time operating data and input the real-time operating data into the dynamic water-solar complementary characteristic model. Use an online learning algorithm to update the parameters of the dynamic water-solar complementary characteristic model and obtain the updated dynamic water-solar complementary characteristic model.

2. The intelligent scheduling method for intelligent mining of water-solar complementary characteristics as described in claim 1, characterized in that, The specific steps of S2.4 include: S2.41: Align the prediction results of photovoltaic power output and the regulation capacity of cascade small hydropower by time point, organize them into a prediction result dataset, and standardize the data in the prediction result dataset. Each row in the prediction result dataset represents the predicted value at a time point. S2.42: Set the number of clusters, and use a clustering algorithm to cluster the data in the standardized prediction result dataset, and output the cluster label for each data point; S2.43: For each cluster, the mean of the data points in each cluster is taken as the cluster center point, and the photovoltaic output and cascade hydropower regulation capacity values ​​of the cluster center point are extracted. If the center point of the cluster shows high photovoltaic output and low small hydropower regulation capacity, it means that under the current circumstances, photovoltaic power is the source of power generation and small hydropower is a supplementary power source. If the center point of the cluster shows low photovoltaic output and high small hydropower regulation capacity, it means that under the current conditions, small hydropower is the source of power generation and photovoltaic power is a supplementary power source. S2.44: Based on the judgment result of S2.43, define the types of complementary modes, including high photovoltaic + low small hydropower and low photovoltaic + high small hydropower; S2.45: Based on the clustering results and complementary patterns, draw a scatter plot of photovoltaic power output and the regulation capacity of cascade small hydropower, and use different colors to represent different clusters. At the same time, mark the center point of each cluster in the scatter plot.

3. The intelligent scheduling method for intelligent mining of water-solar complementary characteristics as described in claim 2, characterized in that, The specific steps of S2.5 include: S2.51: Associate the center point of each cluster with the corresponding meteorological conditions to extract the complementary relationship between photovoltaic and small hydropower under different meteorological conditions; S2.52: Based on the complementary relationship between photovoltaic and small hydropower under different meteorological conditions, a reinforcement learning environment is constructed, including defining a state space, an action space, and a reward function; the state space includes the current meteorological conditions, photovoltaic output, small hydropower regulation capacity, and reservoir water level; the action space includes reservoir water release strategy and photovoltaic output adjustment strategy. S2.53: Load the pre-trained reinforcement learning model and input different meteorological conditions and scheduling strategies into the pre-trained reinforcement learning model to simulate changes in complementary relationships; the scheduling strategies include reservoir water release strategies and photovoltaic power output adjustment strategies. S2.54: Based on the simulation results, extract the variation law of the complementary relationship between photovoltaic and small hydropower under different meteorological conditions and dispatch strategies.

4. The intelligent scheduling method for intelligent mining of water-solar complementary characteristics as described in claim 3, characterized in that, The specific steps of S2.6 include: S2.61: Align the prediction results of the random forest model with the simulation results of reinforcement learning by timestamp and perform feature fusion to obtain a high-dimensional feature dataset; S2.62: Load the pre-built machine learning model and train the pre-built machine learning model using a high-dimensional feature dataset to obtain a dynamic water-photovoltaic complementary characteristic model; S2.63: Obtain meteorological conditions and reservoir water level data at different time scales, and input them into the dynamic water-solar complementary characteristic model to obtain photovoltaic power output and cascade small hydropower regulation capacity at different time scales.

5. The intelligent scheduling method for intelligent mining of water-solar complementary characteristics as described in claim 4, characterized in that, The specific steps of step S3 include: S3.1: Collect photovoltaic power output fluctuation data and cascade small hydropower regulation capacity data; S3.2: Set reservoir scheduling constraints ,in, Indicates the reservoir water level. and These represent the lower and upper limits of the reservoir water level, respectively. Indicates the reservoir flow rate. and These represent the lower and upper limits of the reservoir's flow rate, respectively. S3.3: Setting the multi-objective optimization function ,in, This indicates power generation efficiency, and Cost indicates the cost of generating electricity. Indicates carbon emissions, , , Indicates the weighting coefficient; S3.4: Based on the characteristic model of water-solar complementarity, combined with photovoltaic power output fluctuation data, cascade small hydropower regulation capacity data and reservoir scheduling constraints, a joint power generation model is constructed; S3.5: Use intelligent optimization scheduling algorithm to solve the joint generation model and obtain the optimal scheduling scheme; S3.6: Transmit the optimal scheduling scheme to the joint operation control system.

6. The intelligent scheduling method for intelligent mining of water-solar complementary characteristics as described in claim 5, characterized in that, The specific steps of S3.5 include: S3.51: Set the parameters of the differential evolution algorithm, including population size, scaling factor, crossover probability and maximum number of iterations; S3.52: Randomly generate an initial population, wherein each individual in the population... This represents a scheduling scheme; S3.53: Based on the multi-objective optimization function in S3.3, calculate the objective function value for each individual. ; S3.54: Pass Generate mutated individuals The mutant individuals and the original individuals were cross-operated to generate experimental individuals. ,in, , , This represents three randomly selected distinct individuals. Indicates the scaling factor; S3.55: Comparative test individuals and the original individual The objective function value; like Then select test individuals Moving into the next generation; like Then select the original individual. Moving into the next generation; S3.56: If the current iteration count reaches the maximum iteration count, terminate the differential evolution algorithm and select the individual with the highest objective function value in the population as the optimal scheduling scheme.

7. The intelligent scheduling method for intelligent mining of water-solar complementary characteristics as described in claim 6, characterized in that, The joint operation control system includes grid-connected joint control, fast and smooth control, off-grid coordinated control, and advanced predictive control functions.

8. An intelligent scheduling system for intelligent mining of water-solar complementary characteristics, used to implement the intelligent scheduling method for intelligent mining of water-solar complementary characteristics as described in any one of claims 1-7, characterized in that, include: Data acquisition module, complementary analysis module, joint power generation model construction module, feedback optimization module; The data acquisition module is used to acquire multi-source operation data from distributed photovoltaic power stations and cascade small hydropower systems, perform intelligent preprocessing, and store the preprocessed data in a distributed database. The complementary analysis module uses machine learning algorithms and reinforcement learning techniques to analyze the output characteristics of photovoltaic and small hydropower, explore dynamic complementary relationships, and construct a dynamic hydro-photovoltaic complementary characteristic model. The joint power generation model construction module constructs a joint power generation model based on the water-solar complementary characteristic model, and designs a multi-objective optimization function to solve the optimal scheduling scheme. The feedback optimization module dynamically adjusts the output plan according to the optimal scheduling scheme, monitors the operation effect in real time, and optimizes the model and algorithm through the feedback mechanism.

Images