Photovoltaic-Gas Hybrid Power Supply Control Methods, Devices, Equipment, and Storage Media

Abstract

This application provides a photovoltaic-gas hybrid power supply control method, device, equipment, and storage medium, belonging to the field of new energy power generation technology. The method includes: detecting a system network state switching from online to offline; training a convolutional neural network based on local historical operating data to obtain a second data processing model; using the model parameters of the second data processing model as model parameters of a first data processing model; training the first data processing model based on first operating data obtained in the off-grid state; and obtaining predicted operating data based on the trained first data processing model; adjusting the first control parameters of the photovoltaic power generation module and the second control parameters of the gas-fired power generation module through an optimized control algorithm based on the predicted operating data; and regulating the photovoltaic-gas hybrid power supply system based on the adjusted control parameters. This application can dynamically regulate the system according to actual operating parameters to meet load demands during network interruptions.

Description

Technical Field

[0001] This application belongs to the field of new energy power generation technology, and more specifically, it relates to a photovoltaic-gas hybrid power supply control method, device, equipment, and storage medium. Background Technology

[0002] With the widespread application of renewable energy, hybrid photovoltaic (PV) and gas-fired power supply systems have gradually become a research hotspot in the field of distributed energy due to their ability to comprehensively utilize solar and gas resources, demonstrating significant advantages in ensuring power supply stability and reliability. These hybrid power supply systems typically rely on networks for remote monitoring, optimized scheduling, and interaction with external energy markets.

[0003] However, network failures or communication interruptions are unavoidable problems in actual operation. For example, in scenarios such as industrial parks and remote mountainous areas, when the system network fails or communication is interrupted, existing technologies usually switch to a local preset mode, operate according to fixed parameters, and achieve basic protection only through threshold alarms from local sensors, which cannot adapt to complex operating conditions. Summary of the Invention

[0004] To address the aforementioned issues, this application provides a photovoltaic-gas hybrid power supply control method, device, equipment, and storage medium that can dynamically adjust the system based on actual operating parameters during network interruptions to adapt to complex operating conditions and meet load requirements.

[0005] A first aspect of this application provides a photovoltaic-gas hybrid power supply control method, applied to a photovoltaic-gas hybrid power supply system, the photovoltaic-gas hybrid power supply system including a photovoltaic power generation module and a gas power generation module, the method comprising:

[0006] In response to the detection that the network status of the photovoltaic gas hybrid power supply system has switched from online to offline, the system acquires the first operating data of the photovoltaic gas hybrid power supply system collected within a first time period, and periodically acquires the second operating data of the photovoltaic gas hybrid power supply system after the first time period.

[0007] The second running data is input into the trained first data processing model to obtain the predicted running data; wherein, the first data processing model is obtained by training a convolutional neural network based on local historical running data to obtain the second data processing model, and training the second data processing model based on the first running data to obtain the first data processing model;

[0008] Based on predicted operating data, the first control parameters of the photovoltaic power generation module and the second control parameters of the gas-fired power generation module are adjusted by optimizing the control algorithm.

[0009] The photovoltaic-gas hybrid power supply system is regulated based on the adjusted first and second control parameters.

[0010] A second aspect of this application provides a photovoltaic-gas hybrid power supply control device, applied to a photovoltaic-gas hybrid power supply system, the photovoltaic-gas hybrid power supply system including a photovoltaic power generation module and a gas power generation module, the device comprising:

[0011] The first data acquisition unit is used to acquire the first operating data of the photovoltaic gas-fired hybrid power supply system collected within a first time period in response to the detection that the network status of the photovoltaic gas-fired hybrid power supply system has switched from online to offline, and periodically acquire the second operating data of the photovoltaic gas-fired hybrid power supply system after the first time period.

[0012] The first data processing unit is used to input the second running data into the trained first data processing model to obtain the predicted running data; wherein, the first data processing model is obtained by training a convolutional neural network based on local historical running data to obtain the second data processing model, and training the second data processing model based on the first running data to obtain the first data processing model.

[0013] The first parameter adjustment unit is used to adjust the first control parameter of the photovoltaic power generation module and the second control parameter of the gas-fired power generation module based on the predicted operating data and by optimizing the control algorithm.

[0014] The first control unit is used to regulate the photovoltaic-gas hybrid power supply system based on the adjusted first and second control parameters.

[0015] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above-described photovoltaic-gas hybrid power supply control method.

[0016] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the photovoltaic-gas hybrid power supply control method described above.

[0017] The beneficial effects of the photovoltaic-gas hybrid power supply control method, device, equipment, and storage medium provided in this application embodiment are as follows:

[0018] When the network status switches to offline mode, this embodiment trains a first data processing model adapted to the current state using running data for a first duration. Combined with periodically collected real-time data, it achieves accurate prediction, solving the problem of insufficient prediction accuracy caused by the lack of cloud support in offline mode. The first data processing model is trained based on the parameters of the second data processing model, reducing the system model training time and facilitating timely adjustment of the control parameters of various modules under complex operating conditions to meet load requirements. Furthermore, this embodiment also dynamically adjusts the control parameters of the photovoltaic power generation module and the gas-fired power generation module using optimized control algorithms, enabling rapid response to fluctuations in sunlight and load changes, reducing power supply fluctuations, and ensuring stable load operation.

[0019] Therefore, the embodiments of this application are free from strong dependence on the network in the offline state. In scenarios with unstable networks, such as industrial parks and remote mountainous areas, the system is dynamically adjusted according to the actual operating parameters, which significantly improves the system's anti-interference ability and operational reliability, while reducing energy waste and equipment damage caused by network interruption. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic flowchart of a photovoltaic-gas hybrid power supply control method provided in an embodiment of this application;

[0022] Figure 2 This is a structural block diagram of a photovoltaic-gas hybrid power supply control device provided in an embodiment of this application;

[0023] Figure 3 This is a schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0024] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the following description will be provided in conjunction with the accompanying drawings and specific embodiments.

[0026] A photovoltaic-gas hybrid power supply system is a renewable energy power generation system that combines photovoltaic power generation and gas-fired power generation. This system achieves efficient energy supply by coordinating the operation of photovoltaic and gas-fired power generation modules. For example, when sunlight is insufficient or unstable, the gas-fired power generation modules supplement the power supply, thus ensuring the continuity and stability of the power supply. It is suitable for scenarios with certain requirements for power supply reliability, such as off-grid power supply in remote areas and small-scale industrial and commercial distributed energy systems.

[0027] Photovoltaic power generation modules are used to convert solar energy into electrical energy using the photovoltaic effect. Photovoltaic power generation uses solar panels composed of multiple solar cells containing photovoltaic materials. Multiple photovoltaic modules containing solar panels are connected in series and parallel, together with a supporting structure and an inverter, to form a photovoltaic array.

[0028] Gas-fired power generation modules use combustible gases such as natural gas and biogas as fuel, and generate electricity through an internal combustion engine or gas turbine driving a generator. They are characterized by quick start-up and low pollution. A gas-fired power generation module mainly consists of a gas supply unit, a combustion unit, a generator, and a waste heat recovery unit.

[0029] Figure 1 This is a schematic flowchart of a photovoltaic-gas hybrid power supply control method provided in an embodiment of this application, with reference to... Figure 1 This method is applied to a photovoltaic-gas hybrid power supply system, which includes a photovoltaic power generation module and a gas power generation module. The method may include steps S101 to S104.

[0030] S101: In response to detecting that the network status of the photovoltaic gas hybrid power supply system has switched from online to offline, the system acquires the first operating data of the photovoltaic gas hybrid power supply system collected within a first time period, and periodically acquires the second operating data of the photovoltaic gas hybrid power supply system after the first time period.

[0031] In actual operation, photovoltaic-gas hybrid power supply systems typically maintain connections with external monitoring platforms, dispatch centers, and cloud servers to achieve real-time data uploads and obtain weather information, load change information, and other data. Through network connectivity, the system can receive remote control commands, optimize the coordinated control strategies of the photovoltaic and gas-fired power generation modules, improve energy utilization, and facilitate timely monitoring of the system's operational status by maintenance personnel.

[0032] However, due to factors such as communication environment (e.g., unstable signals in remote areas) and network failures, the system may frequently experience "network offline status," making it unable to interact with external networks or receive remote commands. If the system's network status is detected to have switched from online to offline, the following measures can be taken to ensure that the system dynamically adjusts itself based on actual operating parameters during network interruptions, adapting to complex operating conditions and meeting load requirements.

[0033] When the network is offline, due to the limited local data storage capacity, the original data acquisition frequency needs to be adjusted to a lower first data acquisition frequency (lower than the original frequency) to reduce data storage. When the network is offline, a connection to the cloud server cannot be established, preventing the timely receipt of control commands from the cloud server. In this case, the local data processing model needs to be invoked to adjust the system's operating status in real time to maximize load capacity.

[0034] First, it is necessary to acquire the first operating data of the photovoltaic-gas hybrid power supply system collected within a first time period. After the first time period, the second operating data of the photovoltaic-gas hybrid power supply system is periodically acquired. This can be understood as follows: when the system enters off-grid mode, the first operating data of the photovoltaic-gas hybrid power supply system collected within the first time period is acquired, and this first operating data is used to train a first data processing model, resulting in a trained first data processing model. After the first time period, the second operating data of the system is periodically acquired, and the second operating data of each period is input into the trained first data processing model to obtain the predicted operating data for the next period. Based on this predicted operating data, the control parameters of the photovoltaic-gas hybrid power supply system are adjusted. This embodiment can ensure energy supply to the load when the system enters offline mode, reducing the impact of external factors on the system's operating status.

[0035] S102: Input the second running data into the trained first data processing model to obtain the predicted running data.

[0036] The first data processing model is obtained by training a convolutional neural network based on local historical running data to obtain a second data processing model, and then training the second data processing model based on the first running data to obtain the first data processing model.

[0037] In this embodiment, because training the model requires a significant amount of time and the system's data storage capacity is limited, the frequency of data collection decreases when the system enters offline mode, and the types of data collected are fewer than those in the online state (i.e., the types of data in both the first and second running data are fewer than those in the local historical running data). Therefore, this embodiment first trains the convolutional neural network based on the local historical running data to obtain a more accurate second data processing model; then, it uses the first running data to train the second data processing model to obtain the first data processing model. This model training method reduces training time and increases the accuracy of model predictions. The convolutional neural network in this embodiment can be a Long Short-Term Memory (LSTM) network, a lightweight convolutional network (such as MobileNet), etc.

[0038] After the first data processing model is trained, the periodically obtained second operating data can be input into the trained first data processing model to obtain predicted operating data. Based on the predicted operating data, the control parameters of the photovoltaic gas hybrid power supply system can be adjusted.

[0039] In this embodiment, if the difference between the predicted operating data obtained based on the second operating data acquired in the previous cycle and any type of actual operating data in that cycle is greater than a preset difference threshold, then the parameters of the trained first data processing model are updated based on the second operating data of the system in the current cycle to obtain an updated first data processing model. This embodiment, based on the above method, enables the system to adjust control parameters according to real-time operating data, maintaining normal load operation.

[0040] S103: Based on predicted operating data, the first control parameter of the photovoltaic power generation module and the second control parameter of the gas-fired power generation module are adjusted by optimizing the control algorithm.

[0041] In this embodiment, the photovoltaic-gas hybrid power supply system also includes an energy storage module, which is connected to both the photovoltaic power generation module and the gas power generation module.

[0042] The first control parameters of the photovoltaic power generation module and the second control parameters of the gas-fired power generation module are adjusted by optimizing the control algorithm, including:

[0043] When the output power of the photovoltaic power generation module is greater than the load demand power, the energy storage module is controlled to be in charging mode. The first control parameter of the photovoltaic power generation module is adjusted based on the difference between the output power of the photovoltaic power generation module and the load demand power, and the second control parameter of the gas power generation module is adjusted based on the difference between the output power of the photovoltaic power generation module and the load demand power.

[0044] When the output power of the photovoltaic power generation module is less than or equal to the load demand power, the control energy storage module is in discharge mode, supplying all the output power of the photovoltaic power generation module to the load, and adjusting the second control parameters of the gas power generation module according to the load priority order.

[0045] In this embodiment, when the output power of the photovoltaic (PV) power generation module exceeds the load demand power, the first control parameter of the PV power generation module is adjusted based on the difference between the output power of the PV power generation module and the load demand power. This includes determining the proportion of PV modules in the PV power generation module that charge the energy storage module based on the difference between the output power of the PV power generation module and the load demand power. For example, if the PV power generation module includes 10 PV modules, based on the output power of the PV power generation module, 4 PV modules can be used to charge the energy storage module, and the remaining PV modules can supply power to the load. The second control parameter of the gas-fired power generation module is adjusted based on the difference between the output power of the PV power generation module and the load demand power. For example, the output value of the gas-fired power generation module can be directly turned off, or the output power of the gas-fired power generation module can be adjusted to the minimum output power.

[0046] In this embodiment, when the output power of the photovoltaic power generation module is less than or equal to the load demand power, the second control parameter of the gas power generation module is adjusted according to the load priority order, including: dividing the load into first-level critical load and second-level ordinary load according to the preset priority, and preset the minimum guaranteed power threshold of each level of load; adjusting the initial output power of the gas power generation module to a level that can cover the difference between the output power of the first-level critical load and the photovoltaic power generation module, and controlling the gas valve opening in the second control parameter to increase at a rate of a% / s.

[0047] S104: The photovoltaic-gas hybrid power supply system is regulated based on the adjusted first and second control parameters.

[0048] In this embodiment, the first control parameter corresponds to the operation and regulation parameters of the photovoltaic power generation module (such as the working mode of the photovoltaic inverter, the grouping parameters of the photovoltaic modules, and the photovoltaic output limit threshold, etc.); the second control parameter corresponds to the operation and regulation parameters of the gas power generation module (such as the start / stop threshold of the gas turbine, the power generation regulation coefficient, the fuel supply and the gas valve opening, etc.).

[0049] After obtaining the adjusted first and second control parameters, the system's local master controller will distribute each control parameter to the corresponding execution unit (e.g., the photovoltaic controller of the photovoltaic power generation module and the gas controller of the gas power generation module).

[0050] For photovoltaic power generation modules, their operating status is optimized according to the adjusted first control parameters. For example, when the load is lower than the preset load threshold, the photovoltaic power is limited to avoid energy waste.

[0051] For the gas-fired power generation module, flexible scheduling is achieved based on the adjusted second control parameters. For example, when the photovoltaic output is insufficient, the gas-fired power generation is started to make up for the shortfall according to the power value set by the parameters; when the system load drops sharply, the gas output is reduced or the unit is shut down according to the parameter threshold to reduce fuel consumption.

[0052] There is no strict execution order between S101 and S102. S101 can be executed before S102 or at the same time as S102. Figure 1 The execution method described is merely an example and is not intended to be limiting.

[0053] As can be seen from the above, when the network status switches to offline status, this embodiment trains a first data processing model adapted to the current status using running data for a first duration, and achieves accurate prediction by combining periodically collected real-time data, thus solving the problem of insufficient prediction accuracy caused by the lack of cloud support in offline status. The first data processing model in this embodiment is trained based on the parameters of the second data processing model, reducing the system model training time and facilitating timely adjustment of the control parameters of various modules under complex operating conditions to meet load requirements. Furthermore, this embodiment also dynamically adjusts the control parameters of the photovoltaic power generation module and the gas-fired power generation module using optimized control algorithms, enabling rapid response to fluctuations in sunlight and load changes, reducing power supply fluctuations, and ensuring stable load operation. Therefore, this embodiment eliminates strong dependence on the network in offline status, significantly improving the system's anti-interference capability and operational reliability in scenarios with unstable networks, such as industrial parks and remote mountainous areas, while reducing energy waste and equipment wear caused by network interruptions.

[0054] In one embodiment of this application, the photovoltaic power generation module includes a photovoltaic controller, the gas power generation module includes a gas controller, and the photovoltaic controller is connected to the gas controller;

[0055] The photovoltaic-gas hybrid power supply system is regulated based on the adjusted first and second control parameters, including:

[0056] The adjusted first control parameter is sent to the photovoltaic controller, and the adjusted second control parameter is sent to the gas controller;

[0057] In response to the photovoltaic controller not receiving the first control parameter, the first control parameter is forwarded to the photovoltaic controller through the gas controller;

[0058] The photovoltaic-gas hybrid power supply system is regulated based on the adjusted first and second control parameters.

[0059] In this embodiment, the normal communication process is as follows: the system directly sends the adjusted first control parameter to the photovoltaic controller and the second control parameter directly to the gas controller, realizing independent control of the two modules. However, the communication of the photovoltaic controller is easily affected by the environment. When the photovoltaic controller fails to receive the first control parameter due to a communication link failure (such as line interference or loose interface), the system automatically triggers a backup path, that is, the first control parameter is forwarded from the photovoltaic controller to the photovoltaic controller via the gas controller through a preset connection between the photovoltaic controller and the gas controller (such as RS485 bus or wireless Mesh network).

[0060] This embodiment increases the connection between the photovoltaic controller and the gas controller, so that the first control parameter can be received regardless of whether a communication link failure occurs between the photovoltaic controller and the system's main controller. Ultimately, this ensures that both the photovoltaic controller and the gas controller obtain the corresponding control parameters, and the two work together based on their respective control parameters (such as timely supplementary energy from gas generation when photovoltaic power generation is insufficient), thereby achieving stable system operation.

[0061] In one embodiment of this application, the photovoltaic-gas hybrid power supply control method further includes:

[0062] In response to the detection that the network status of the photovoltaic gas hybrid power supply system has switched from offline to online, the dataset collected during the offline period and the model parameters of the first data processing model are sent to the cloud server.

[0063] Receive the residual correction model sent by the cloud server, and simulate the voltage change after the parameter update of the first data processing model through simulation test based on the residual correction model;

[0064] If the voltage change is less than the preset threshold, the parameters of the first data processing model are updated according to the step update strategy to obtain the target data processing model.

[0065] The target control parameters are obtained by processing the periodically acquired operating data of photovoltaic power generation modules and gas power generation modules based on the target data processing model, and the photovoltaic gas hybrid power supply system is regulated based on the target control parameters.

[0066] In this embodiment, when the system network fails or communication is interrupted, it not only affects the system's transition from online to offline status but also its recovery from offline to online status. During the network outage, the system's main controller may experience data deviations (such as sensor drift), causing the parameters of the first data processing model to deviate from the cloud benchmark (e.g., the PID control parameters of the gas-fired power generation module). Directly overwriting local parameters after network restoration may cause system oscillations. Therefore, after network restoration, the system needs to update the parameters of the first data processing model based on residual fusion using federated learning.

[0067] In response to the detection that the network status has switched from offline to online, the dataset collected during the offline period and the model parameters of the first data processing model are sent to the cloud server, including:

[0068] The dataset collected offline and the model parameters of the first data processing model are subjected to integrity verification, timestamp continuity verification, and sensor drift compensation verification to generate a standardized dataset, which is then sent to a cloud server. Integrity verification may involve confirming whether key control parameters have been tampered with through hash value comparison; timestamp continuity verification may involve identifying data recording interruptions and marking the missing data type; and sensor drift compensation verification may involve correcting offset data based on a preset sensor error model. This dataset includes, but is not limited to, hourly power generation, conversion efficiency, and temperature coefficient of photovoltaic power generation modules, and gas consumption, unit speed, cylinder temperature, and load-side power fluctuation curves and harmonic content of gas-fired power generation modules.

[0069] After receiving the standardized dataset, the cloud server analyzes the global data (including operational data from other systems and meteorological forecast data) to generate relevant instructions and models (e.g., residual correction models) for optimizing the first data processing model, and then encrypts and distributes the residual correction model to the local central controller.

[0070] After receiving the residual correction model, the local controller simulates the voltage change after the parameter update of the first data processing model based on the residual correction model through simulation test. This can be represented as the voltage change after the parameter update of the first data processing model is obtained by reproducing the system operating environment through the digital twin simulation module and adding environmental parameters (e.g., ±5% illumination, gas pressure disturbance parameters, etc.).

[0071] If the voltage change is less than a preset threshold, the parameters of the first data processing model are updated according to a tiered update strategy to obtain the target data processing model. If the voltage change is greater than or less than the preset threshold, a multi-level risk control mechanism is activated to promptly synchronize the risk to maintenance personnel, preventing parameter updates from impacting the system. The process of updating the parameters of the first data processing model according to the tiered update strategy to obtain the target data processing model includes:

[0072] The parameter correction amount corresponding to the residual correction model is divided into multiple step intervals, and each step interval corresponds to a different parameter update weight.

[0073] The target step interval for parameter correction of the first data processing model is determined based on the error distribution of the running data within the offline time period.

[0074] The parameters of the first data processing model are updated in stages according to the update weights corresponding to the target step intervals. After each update stage, the updated parameters of the first data processing model are verified by the validation set to obtain the verification results.

[0075] If the verification results meet the preset accuracy requirements, the next update stage will begin, continuing until all parameter updates for all step intervals are completed, and the target data processing model is obtained.

[0076] In this embodiment, the parameter correction amount corresponding to the residual correction model is divided into multiple stepped intervals, for example, the total correction amount is divided into 3 stepped intervals:

[0077] Step interval 1: 30% of the correction amount (U_ref+6V, K_p+0.12), update weight 30%;

[0078] Step interval 2: 50% of the correction amount (U_ref+10V, K_p+0.2), update weight 50%;

[0079] Step interval 3: 20% of the correction amount (U_ref+4V, K_p+0.08), update weight 20%.

[0080] Where U_ref represents the photovoltaic MPPT voltage reference value, and K_p represents the power regulation coefficient of the gas turbine unit in the gas-fired power generation module.

[0081] Analysis of the error distribution of operational data during the offline period revealed that the prediction error of photovoltaic power generation was mainly concentrated in ±8%, while the power regulation coefficient of the gas turbine unit reached ±15%. According to the error matching principle, the step interval corresponding to the parameter with larger error is updated first, so the target step interval for the first stage is "step 2".

[0082] The parameters of the first data processing model are updated in stages according to the update weights corresponding to the target tier intervals of the first stage. After each update stage, the updated parameters of the first data processing model are validated using a validation set to obtain the validation results. This validation set includes operating condition data under historical extreme weather conditions and operating condition data under conditions of sudden changes in load power.

[0083] If the verification results meet the preset accuracy requirements, for example, if the final verification error is ±3% and ±6%, which meets the accuracy requirements (≤±10%), then the target data processing model is obtained.

[0084] In this embodiment, if the verification result does not meet the preset accuracy requirements, the model parameter update is paused, the error distribution is redefined, or the stepped interval is divided into multiple sub-steps, and the verification is repeated until the parameter updates for all stepped intervals are completed, thus obtaining the target data processing model. Based on the target data processing model, the periodically acquired operating data of the photovoltaic power generation module and the gas-fired power generation module are processed to obtain the target control parameters, and the photovoltaic-gas hybrid power supply system is regulated based on the target control parameters.

[0085] This embodiment dynamically selects the update interval by matching the error distribution and combines the adjustment rhythm of the control parameters in the stage verification. This can avoid system fluctuations caused by sudden parameter changes and improve model accuracy through precise correction. While ensuring stable power supply, it can achieve precise collaboration between local and cloud models and improve the operational reliability and control accuracy of the hybrid power supply system after it is restored from offline to online.

[0086] In one embodiment of this application, the parameter correction amount corresponding to the residual correction model is divided into multiple stepped intervals, including:

[0087] The distribution characteristics of the corrected data based on the historical parameters of the first data processing model are used to determine the threshold for dividing the stepped intervals. The threshold for dividing the intervals is positively correlated with the load fluctuation coefficient of the photovoltaic-gas hybrid power supply system.

[0088] The parameter correction amount corresponding to the residual correction model is divided into multiple stepped intervals based on the threshold of the stepped interval.

[0089] When the load fluctuation coefficient is greater than the preset fluctuation threshold, the number of step intervals is increased based on the first step length.

[0090] In this embodiment, the parameter correction amount corresponding to the residual correction model refers to the specific numerical change in the residual correction model generated by the cloud server, used to adjust the parameters of the first data processing model. For example, if the photovoltaic output prediction coefficient of the first data processing model needs to be corrected from 0.7 to 0.9, the parameter correction amount is +0.2, reflecting the deviation between the local model and the cloud-optimized model. The distribution characteristics of historical parameter correction data are the statistical patterns of the parameter correction amounts of the first data processing model during the historical operation of the system, including the concentration range, frequency distribution, or probability of extreme values. For example, if historical data shows that 80% of the correction amounts are concentrated within ±0.1, the core range for dividing the stepped intervals can be determined accordingly. The division threshold is the critical value used to divide the stepped intervals, determining the start and end boundaries of each stepped interval. For example, when the division threshold is set to 0.05 and 0.15, the correction amount of +0.2 can be divided into three stepped intervals.

[0091] The load fluctuation coefficient is a quantitative indicator describing the severity of load power changes in a photovoltaic-gas hybrid power supply system. It is typically expressed as the ratio of the maximum change in load power per unit time to the rated power. For example, if the load power suddenly increases from 50kW to 80kW within 10 minutes, and the rated power is 100kW, the load fluctuation coefficient is 30%. The preset fluctuation threshold is a pre-defined critical value for the load fluctuation coefficient, used to determine if the load is experiencing severe fluctuations. When the actual load fluctuation coefficient exceeds this value, an adjustment mechanism for the number of stepped intervals is triggered. The first step size is a fixed increment that increases the number of stepped intervals when the load fluctuation coefficient exceeds the preset threshold (e.g., a first step size of 2 increases the number of stepped intervals from 3 to 5). The step size is set based on the system's tolerance to load fluctuations, ensuring that the level of interval refinement matches the intensity of the fluctuations.

[0092] In this embodiment, the division threshold is determined based on the distribution characteristics of historical parameter correction data, ensuring that the stepped intervals closely match the actual system conditions. Furthermore, the division threshold is positively correlated with the load fluctuation coefficient; the threshold changes accordingly when load fluctuations are large. Once the preset threshold is exceeded, the number of stepped intervals is increased by the first step length. This interval adjustment method in this embodiment reduces the impact of sudden parameter changes, adapts to different load conditions, and enhances system stability while ensuring model optimization efficiency.

[0093] In one embodiment of this application, the updated parameters of the first data processing model are validated using a validation set to obtain a validation result, including:

[0094] The validation set is input into the updated first data processing model, and the mean absolute error and root mean square error of the output results compared with the standard results are calculated.

[0095] The verification results are obtained based on the comparison between the mean absolute error and the first error threshold, and the comparison between the root mean square error and the second error threshold.

[0096] In this embodiment, if the mean absolute error is less than the first error threshold and the root mean square error is less than the second error threshold, the verification result is determined to meet the preset accuracy requirement; otherwise, the verification result is determined not to meet the preset accuracy requirement.

[0097] This embodiment verifies model parameters using both mean absolute error (MAE) and root mean square error (RMSE), enabling a comprehensive evaluation of the updated model's accuracy. MAE reflects overall deviation, while RMSE amplifies the impact of extreme deviations; combining the two allows for accurate assessment of model performance. A dual-threshold determination mechanism ensures the quality of model updates, avoids misjudgments based on a single indicator, effectively guarantees the reliability of the updated model, and provides precise control data for stable system operation.

[0098] Corresponding to the photovoltaic-gas hybrid power supply control method in the above embodiments, Figure 2This is a structural block diagram of a photovoltaic-gas hybrid power supply control device according to an embodiment of this application. For ease of explanation, only the parts relevant to the embodiment of this application are shown. References Figure 2 The photovoltaic-gas hybrid power supply control device 20 is applied to a photovoltaic-gas hybrid power supply system, which includes a photovoltaic power generation module and a gas power generation module. The device includes: a first data acquisition unit 21, a first data processing unit 22, a first parameter adjustment unit 23, and a first control unit 24.

[0099] The first data acquisition unit 21 is used to acquire the first operating data of the photovoltaic gas hybrid power supply system collected within a first time period in response to the detection that the network status of the photovoltaic gas hybrid power supply system has switched from online to offline, and periodically acquire the second operating data of the photovoltaic gas hybrid power supply system after the first time period.

[0100] The first data processing unit 22 is used to input the second running data into the trained first data processing model to obtain the predicted running data; wherein, the first data processing model is obtained by training a convolutional neural network based on local historical running data to obtain the second data processing model, and training the second data processing model based on the first running data to obtain the first data processing model;

[0101] The first parameter adjustment unit 23 is used to adjust the first control parameter of the photovoltaic power generation module and the second control parameter of the gas power generation module based on the predicted operating data and by optimizing the control algorithm.

[0102] The first control unit 24 is used to regulate the photovoltaic-gas hybrid power supply system based on the adjusted first and second control parameters.

[0103] In one embodiment of this application, the photovoltaic power generation module includes a photovoltaic controller, the gas power generation module includes a gas controller, and the photovoltaic controller is connected to the gas controller.

[0104] The first control unit 24 is specifically used to send the adjusted first control parameters to the photovoltaic controller and the adjusted second control parameters to the gas controller;

[0105] In response to the photovoltaic controller not receiving the first control parameter, the first control parameter is forwarded to the photovoltaic controller through the gas controller;

[0106] The photovoltaic-gas hybrid power supply system is regulated based on the adjusted first and second control parameters.

[0107] In one embodiment of this application, the photovoltaic-gas hybrid power supply control device 20 further includes:

[0108] The second data acquisition unit is used to send the dataset collected during the offline period and the model parameters of the first data processing model to the cloud server in response to the detection that the network status of the photovoltaic gas hybrid power supply system has switched from offline to online.

[0109] The second data processing unit is used to receive the residual correction model sent by the cloud server, and simulate the voltage change after the parameter update of the first data processing model through simulation test based on the residual correction model.

[0110] The second parameter adjustment unit is used to update the parameters of the first data processing model according to the step update strategy to obtain the target data processing model if the voltage change is less than the preset threshold.

[0111] The second control unit is used to process the periodically acquired operating data of the photovoltaic power generation module and the gas power generation module based on the target data processing model to obtain target control parameters, and to regulate the photovoltaic gas hybrid power supply system based on the target control parameters.

[0112] In one embodiment of this application, the second parameter adjustment unit is specifically used for:

[0113] The parameter correction amount corresponding to the residual correction model is divided into multiple step intervals, and each step interval corresponds to a different parameter update weight.

[0114] The target step interval for parameter correction of the first data processing model is determined based on the error distribution of the running data within the offline time period.

[0115] The parameters of the first data processing model are updated in stages according to the update weights corresponding to the target step intervals. After each update stage, the updated parameters of the first data processing model are verified by the validation set to obtain the verification results.

[0116] If the verification results meet the preset accuracy requirements, the next update stage will begin, continuing until all parameter updates for all step intervals are completed, and the target data processing model is obtained.

[0117] In one embodiment of this application, the second parameter adjustment unit is specifically used for:

[0118] The distribution characteristics of the corrected data based on the historical parameters of the first data processing model are used to determine the threshold for dividing the stepped intervals. The threshold for dividing the intervals is positively correlated with the load fluctuation coefficient of the photovoltaic-gas hybrid power supply system.

[0119] The parameter correction amount corresponding to the residual correction model is divided into multiple stepped intervals based on the threshold of the stepped interval.

[0120] In one embodiment of this application, the second parameter adjustment unit is further configured to:

[0121] When the load fluctuation coefficient is greater than the preset fluctuation threshold, the number of step intervals is increased based on the first step length.

[0122] In one embodiment of this application, the second parameter adjustment unit is specifically used for:

[0123] The validation set is input into the updated first data processing model, and the mean absolute error and root mean square error of the output results compared with the standard results are calculated.

[0124] The verification results are obtained based on the comparison between the mean absolute error and the first error threshold, and the comparison between the root mean square error and the second error threshold.

[0125] See Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of the units in the aforementioned device embodiments, for example... Figure 2 The functions of the first data acquisition unit 21, the first data processing unit 22, the first parameter adjustment unit 23, and the first control unit 24 are shown.

[0126] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0127] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0128] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory.

[0129] In specific implementations, the processor 301, input device 302, and output device 303 described in the embodiments of this application can execute the implementation method described in the photovoltaic-gas hybrid power supply control method provided in the embodiments of this application, or they can execute the implementation method of the electronic device described in the embodiments of this application, which will not be repeated here.

[0130] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0131] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the foregoing embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device. Furthermore, the computer-readable storage medium can include both internal and external storage units of the electronic device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0132] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0133] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the electronic devices and units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0134] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces or units, or it may be an electrical, mechanical, or other form of connection.

[0135] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0136] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0137] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A photovoltaic-gas hybrid power supply control method, characterized in that, Applied to a photovoltaic-gas hybrid power supply system, the photovoltaic-gas hybrid power supply system including a photovoltaic power generation module and a gas power generation module, the method includes: In response to the detection that the network status of the photovoltaic gas hybrid power supply system has switched from online to offline, the system acquires the first operating data of the photovoltaic gas hybrid power supply system collected within a first time period, and periodically acquires the second operating data of the photovoltaic gas hybrid power supply system after the first time period. The second running data is input into the trained first data processing model to obtain the predicted running data; wherein, the first data processing model is obtained by training a convolutional neural network based on local historical running data to obtain a second data processing model, and training the second data processing model based on the first running data to obtain the first data processing model; Based on the predicted operating data, the first control parameter of the photovoltaic power generation module and the second control parameter of the gas-fired power generation module are adjusted by optimizing the control algorithm; The photovoltaic-gas hybrid power supply system is regulated based on the adjusted first and second control parameters; The method further includes: In response to the detection that the network status of the photovoltaic gas hybrid power supply system has switched from offline to online, the dataset collected during the offline period and the model parameters of the first data processing model are sent to the cloud server. Receive the residual correction model sent by the cloud server, and simulate the voltage change after the parameter update of the first data processing model through simulation test based on the residual correction model; If the voltage change is less than a preset threshold, the parameters of the first data processing model are updated according to the step update strategy to obtain the target data processing model. The target control parameters are obtained by processing the periodically acquired operating data of the photovoltaic power generation module and the gas power generation module based on the target data processing model, and the photovoltaic gas hybrid power supply system is regulated based on the target control parameters.

2. The method as described in claim 1, characterized in that, The photovoltaic power generation module includes a photovoltaic controller, the gas power generation module includes a gas controller, and the photovoltaic controller is connected to the gas controller. The regulation of the photovoltaic-gas hybrid power supply system based on the adjusted first and second control parameters includes: The adjusted first control parameter is sent to the photovoltaic controller, and the adjusted second control parameter is sent to the gas controller; In response to the photovoltaic controller not receiving the first control parameter, the first control parameter is forwarded to the photovoltaic controller through the gas controller; The photovoltaic-gas hybrid power supply system is regulated based on the adjusted first and second control parameters.

3. The method as described in claim 1, characterized in that, The step of updating the parameters of the first data processing model according to the step update strategy to obtain the target data processing model includes: The parameter correction amount corresponding to the residual correction model is divided into multiple step intervals, and each step interval corresponds to a different parameter update weight. The target step interval for parameter correction of the first data processing model is determined based on the error distribution of the running data within the offline time period. The parameters of the first data processing model are updated in stages according to the update weights corresponding to the target step intervals. After each update stage, the updated parameters of the first data processing model are verified by a validation set to obtain the verification results. If the verification result meets the preset accuracy requirements, then proceed to the next update stage until the parameter updates for all step intervals are completed, and the target data processing model is obtained.

4. The method as described in claim 3, characterized in that, The step of dividing the parameter correction amount corresponding to the residual correction model into multiple stepped intervals includes: The distribution characteristics of the corrected data based on the historical parameters of the first data processing model are used to determine the threshold for dividing the stepped intervals. The threshold for dividing the intervals is positively correlated with the load fluctuation coefficient of the photovoltaic-gas hybrid power supply system. Based on the threshold for dividing the stepped intervals, the parameter correction amount corresponding to the residual correction model is divided into multiple stepped intervals.

5. The method as described in claim 4, characterized in that, The method further includes: When the load fluctuation coefficient is greater than the preset fluctuation threshold, the number of step intervals is increased based on the first step length.

6. The method as described in claim 3, characterized in that, The step of validating the updated parameters of the first data processing model using a validation set to obtain validation results includes: The validation set is input into the updated first data processing model, and the mean absolute error and root mean square error of the output result compared with the standard result are calculated. The verification result is obtained based on the comparison result of the mean absolute error and the first error threshold, and the comparison result of the root mean square error and the second error threshold.

7. A photovoltaic-gas hybrid power supply control device, characterized in that, An application in a photovoltaic-gas hybrid power supply system, the photovoltaic-gas hybrid power supply system comprising a photovoltaic power generation module and a gas power generation module, the device comprising: The first data acquisition unit is used to acquire the first operating data of the photovoltaic gas-fired hybrid power supply system collected within a first time period in response to the detection that the network status of the photovoltaic gas-fired hybrid power supply system has switched from online to offline, and periodically acquire the second operating data of the photovoltaic gas-fired hybrid power supply system after the first time period. The first data processing unit is used to input the second running data into a trained first data processing model to obtain predicted running data; wherein, the first data processing model is obtained by training a convolutional neural network based on local historical running data to obtain a second data processing model, and training the second data processing model based on the first running data to obtain the first data processing model; The first parameter adjustment unit is used to adjust the first control parameter of the photovoltaic power generation module and the second control parameter of the gas power generation module based on the predicted operating data and through an optimized control algorithm. The first control unit is used to regulate the photovoltaic-gas hybrid power supply system based on the adjusted first control parameters and the second control parameters; The device further includes: The second data acquisition unit is used to send the dataset collected during the offline period and the model parameters of the first data processing model to the cloud server in response to the detection that the network status of the photovoltaic gas hybrid power supply system has switched from offline to online. The second data processing unit is used to receive the residual correction model sent by the cloud server, and simulate the voltage change after the parameter update of the first data processing model through simulation test based on the residual correction model. The second parameter adjustment unit is used to update the parameters of the first data processing model according to the step update strategy to obtain the target data processing model if the voltage change is less than the preset threshold. The second control unit is used to process the periodically acquired operating data of the photovoltaic power generation module and the gas power generation module based on the target data processing model to obtain target control parameters, and to regulate the photovoltaic gas hybrid power supply system based on the target control parameters.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 6.

Images