Photovoltaic power generation control method and system for breeding power supply

Abstract

The invention relates to the technical field of photovoltaic power generation control, and discloses a photovoltaic power generation control method and system for power supply for cultivation, and the method comprises the steps: collecting the current power state information; executing path selection and load priority mapping; generating an energy distribution plan of photovoltaic power generation through a prediction enhanced energy feed-forward regulation mechanism; executing power supply switching control; the feedback control process of photovoltaic power generation is optimized; and outputting the optimized power supply switching instruction set. In the prior art, there is no strategy for photovoltaic power generation residual power utilization and dynamic switching precision control, and especially under the condition that commercial power in a culture area is frequently interrupted or the voltage is unstable, self-adaptive closed-loop control of photovoltaic priority and commercial power supplement cannot be realized. The photovoltaic energy utilization efficiency of the aquaculture power utilization system is improved by introducing state driving path selection, dynamic energy distribution and a fuzzy feedback correction mechanism.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photovoltaic power generation control, and particularly relates to a photovoltaic power generation control method and system for power supply in breeding. BACKGROUND

[0002] At present, in rural breeding farms, electric aquaculture systems and remote ranches and other scenes, due to remote location or insufficient extension capacity of terminal power grid lines, the problems of unstable voltage, frequent power outages or limited power supply time often occur. In order to improve the reliability of power supply, more and more breeders introduce distributed photovoltaic systems as auxiliary energy, expecting to reduce electricity costs and alleviate power outages through the mode of photovoltaic priority and power supply. However, the existing power supply switching control method mainly has the following deficiencies: (1) The traditional photovoltaic and power supply hybrid system usually presets a simple switching logic (such as enabling photovoltaic when the photovoltaic voltage is higher than a certain threshold, and vice versa), which fails to dynamically adjust the priority according to the real-time changes of the load, the power fluctuation state, the weather trend and other multi-source information, resulting in low utilization rate of photovoltaic power generation and unnecessary increase of power consumption. (2) Some control systems use voltage threshold or timing switching methods without introducing prediction and judgment of the remaining power of photovoltaic and dynamic perception of the load, which leads to a lag in the response of the switching strategy when the light suddenly weakens or the load suddenly changes, and easily causes power interruption, equipment damage or other problems. (3) It cannot support multi-source path and load cascade mapping scheduling: in the actual breeding scene, the power supply path often includes not only photovoltaic and power supply, but also battery energy storage or diesel backup power generation equipment, and different loads have different requirements for power supply quality (such as oxygenation pumps that require priority power supply, and lighting systems that can be delayed).

[0003] Therefore, there is an urgent need for a photovoltaic power generation control method for power supply in breeding, which maximizes the utilization efficiency of photovoltaic power generation under the premise of ensuring stable operation of the load, dynamically coordinates the matching relationship between power supply and photovoltaic load capacity, and improves the overall power supply adaptability, safety and operation economy in the breeding scene. SUMMARY

[0004] In view of the above technical deficiencies, the purpose of the present application is to provide a photovoltaic power generation control method for power supply in breeding, which aims to solve the technical problems that the prior art lacks strategies for utilizing the remaining power of photovoltaic power generation and dynamically controlling the switching precision, especially under the condition of frequent power outages or unstable voltage in the breeding area, and cannot realize adaptive closed-loop control of "photovoltaic priority and power supply".

[0005] To solve the above technical problems, the present application adopts the following technical scheme: the present application provides a photovoltaic power generation control method for power supply in breeding,

[0006] The photovoltaic power generation control method for power supply in breeding comprises:

[0007] Step S10: Collecting current state information by the STM32 controller, inputting the current state information into a preset power supply mode recognition model, and outputting a current power supply mode state identifier by the power supply mode recognition model ;

[0008] Step S20: Based on the current power supply mode state identifier Adopting a state-aware driving scheduling mechanism to perform a path selection task and a load priority mapping task, and outputting a dynamic energy routing matrix ;

[0009] Step S30: According to the dynamic energy routing matrix Adopting a prediction-enhanced energy feedforward regulation mechanism to perform decision control of photovoltaic power generation, and outputting an energy distribution plan matrix

[0010] Step S40: Based on the energy distribution plan matrix, adopting a multi-level threshold gating and state inversion joint mechanism to perform power supply switching control processing, and outputting a power supply switching instruction set

[0011] Step S50: Applying the power supply switching instruction set, and adopting a fuzzy correction method based on residual power dynamic feedback to perform feedback control of photovoltaic power generation, and outputting an optimized power supply switching instruction set.

[0012] Preferably, in step S10, the current state information includes current photovoltaic power generation , battery state of charge , mains voltage , mains frequency , and historical light trend data ; the current power supply mode state identifier includes a photovoltaic power generation priority mode, a battery priority mode, a mains power supplement mode, and an off-grid power supply mode; the power supply mode recognition model is constructed based on a multi-layer perception network and a gated recurrent unit network, including an input layer for receiving current state information from the STM32 controller; a feature fusion layer for extracting composite features in the current state information using first-layer fully connected neurons ReLU and second-layer fully connected neurons LeakyReLU; a time series perception layer for extracting historical light fluctuation features, SOC fall curve features, and mains fluctuation period features in the composite features using a gated recurrent large network unit; a state mapping layer for mapping the historical light fluctuation features, the SOC fall curve features, and the mains fluctuation period features in the composite features into a fixed-dimensional state space output and a confidence level using a softmax classification function; and an output layer for outputting the current power supply mode state identifier based on the state space output and the confidence level

[0013] Preferably, in step S20, based on the current power supply mode state identifier The state-aware driving scheduling mechanism is adopted to execute the path selection task and the load priority mapping task, and output a dynamic energy routing matrix The step specifically comprises:

[0014] Step S201: First, based on the current power supply mode state identifier Execute the path selection task:

[0015] If the current power supply mode state identifier is the photovoltaic power generation priority mode, activate the first path, and the first path is represented as ;

[0016] If the current power supply mode state identifier is the battery priority mode, activate the second path, and the second path is represented as ;

[0017] If the current power supply mode state identifier is the city power supply mode, activate the third path, and the third path is represented as ;

[0018] If the current power supply mode state identifier is the off-grid power supply mode, activate the fourth path, and the fourth path is represented as ;

[0019] Wherein, represents the photovoltaic power supply; represents the battery discharge path; represents the city power supply; represents the high-priority load; represents the battery charging path; represents the low-priority load; represents that the photovoltaic power supply directly supplies power to the first-priority load; represents that the photovoltaic power supply is preferentially used for battery charging; represents that the battery discharges to supply power to the first-priority load; represents that the battery discharges to supply power to the low-priority load; represents that the city power directly supplies power to the first-priority load; represents that the city power directly supplies power to the low-priority load; represents that the battery independently supplies power to the critical load under off-grid conditions;

[0020] Step S202: re-performing the load priority mapping task: obtaining the current load power, constructing the path activation priority rule based on the current load power; performing dynamic path priority mapping processing on the first path, the second path, the third path and the fourth path using the path activation priority rule, and finally outputting a dynamic energy routing matrix .

[0021] Preferably, in step S30, the dynamic energy routing matrix The step of outputting the energy distribution plan matrix by performing decision control of photovoltaic power generation using the prediction-enhanced energy feedforward adjustment mechanism specifically includes:

[0022] Step S301: obtaining the paths of the photovoltaic power supply contained in the dynamic energy routing matrix , and constructing a photovoltaic energy supply path; collecting the current photovoltaic power generation power in the photovoltaic energy supply path, and if the current photovoltaic power generation power is all greater than the preset activation load power requirement, marking the photovoltaic energy supply path as a selectable energy path;

[0023] Step S302: based on the selectable energy path, performing photovoltaic power generation prediction calculation using a piecewise weighted trend fitting method to output the next moment photovoltaic prediction power;

[0024] Step S303: when the photovoltaic prediction power is less than the preset photovoltaic power generation threshold, setting the battery charging current of the battery charging path in the selectable energy path to the preset maximum charging current; when the photovoltaic prediction power is greater than or equal to the preset photovoltaic power generation threshold, setting the battery charging current of the battery charging path in the selectable energy path to 0.5 times the preset maximum charging current; finally outputting an energy distribution plan matrix, which is used to indicate the power distribution of the photovoltaic power supply, the battery power supply and the mains power supply to each load path.

[0025] Preferably, in step S30, based on the selectable energy path, the step of performing photovoltaic power generation prediction calculation using a piecewise weighted trend fitting method to output the next moment photovoltaic prediction power specifically includes:

[0026] Obtaining historical photovoltaic prediction power data, extracting short-period photovoltaic power generation change characteristics based on the historical photovoltaic prediction power data using a short sliding window gradient extraction principle; extracting long-period photovoltaic power generation change characteristics based on the historical photovoltaic prediction power data using a long sliding window linear fitting principle; constructing a short-period change rate parameter and a long-period trend slope parameter based on the short-period photovoltaic power generation change characteristics and the long-period photovoltaic power generation change characteristics, respectively;

[0027] Based on the short-period change rate parameter and the long-period trend slope parameter, a weighted prediction function is constructed using a weighted nonlinear combination fitting method, and the weighted prediction function outputs a preliminary photovoltaic prediction power;

[0028] The offset confidence constraint method is used to suppress errors of the preliminary photovoltaic predicted power, and a final photovoltaic predicted power at the next moment is output.

[0029] Preferably, in step S40, the multi-level threshold gating and state inversion joint mechanism is used based on the energy distribution plan matrix to perform power supply switching control processing, and the step of outputting the power supply switching instruction set comprises the following steps:

[0030] Step S401: Extract the target load set and the current resource state from the energy distribution plan matrix, and introduce a multi-level dynamic judgment threshold, which includes a photovoltaic available power upper threshold , a photovoltaic available power lower threshold , a mains voltage stability upper threshold , a mains voltage stability lower threshold , and a battery SOC standby threshold ; based on the multi-level dynamic judgment threshold, the power supply path enable state in the energy distribution plan matrix is judged in real time, and a path candidate state set with a normal power supply path enable state is output.

[0031] Step S402: Based on the path candidate state set, a steady-state probability evaluation mechanism based on a Markov state transition graph is used to execute path switching decision judgment, and a power supply path confirmation set is output.

[0032] Step S403: According to the power supply path confirmation set, a preset finite state machine FSM is used to dynamically generate a power supply switching instruction set, which is used to control the circuit breaker and the relay to execute path start-stop operation; the state of the finite state machine FSM is defined as photovoltaic power supply state, mains power supply state, battery power supply state and hybrid power supply state.

[0033] Preferably, in step S50, the power supply switching instruction set is applied, and a fuzzy correction method based on residual power dynamic feedback is used to execute feedback control of photovoltaic power generation, and the step of outputting an optimized power supply switching instruction set comprises the following steps:

[0034] Step S501: Collect the actual load consumption power and the actual photovoltaic output power after applying the power supply switching instruction set, and calculate the photovoltaic residual adjustable power based on the actual load consumption power and the actual photovoltaic output power; if , it indicates that the photovoltaic power supply is insufficient, and the mains or the battery needs to be supplemented; if , it indicates that there is still photovoltaic redundancy.

[0035] Step S502: Based on the photovoltaic residual adjustable power A dynamic fluctuation adjustment factor is constructed, and a fuzzy rule controller is used to perform sensitivity correction operations on the power supply path switching strategy based on the dynamic fluctuation adjustment factor, and outputs a set of control adjustment factors; wherein, the set of control adjustment factors is used to adjust the trigger boundary and waiting time threshold of path switching;

[0036] Step S503: Perform fuzzy correction operation on the power supply switching instruction set according to the control adjustment factor set, including dynamic delay triggering, path priority adjustment and boundary gate value correction processing, and finally output an optimized power supply switching instruction set.

[0037] The present invention also provides a photovoltaic power generation control system for power supply in aquaculture, comprising:

[0038] The power supply status identification module is used to collect current status information through the STM32 controller, input the current status information into a preset power supply mode identification model, and the power supply mode identification model outputs the current power supply mode status identifier. ;

[0039] The path and priority scheduling module is used to schedule based on the current power supply mode status identifier. A state-aware, driven scheduling mechanism is used to execute path selection and load priority mapping tasks, outputting a dynamic energy routing matrix. ;

[0040] The energy feedforward control module is used to adjust the dynamic energy routing matrix. A predictive-enhanced energy feedforward regulation mechanism is used to implement decision control for photovoltaic power generation, and an energy allocation plan matrix is ​​output.

[0041] The power supply switching control module is used to perform power supply switching control processing based on the energy allocation plan matrix using a multi-level threshold gating and state inversion joint mechanism, and outputs a power supply switching instruction set.

[0042] A feedback correction control module is used to apply a power supply switching instruction set and employ a fuzzy correction method based on dynamic feedback of remaining power to execute feedback control of photovoltaic power generation, outputting an optimized power supply switching instruction set. This invention also provides a photovoltaic power generation control device for aquaculture, comprising: a memory, a processor, and a photovoltaic power generation control program for aquaculture stored in the memory and executable on the processor. When executed by the processor, the photovoltaic power generation control program for aquaculture implements a photovoltaic power generation control method for aquaculture.

[0043] The present invention also provides a computer program product, including a photovoltaic power generation control program for power supply in aquaculture, wherein the photovoltaic power generation control program for power supply in aquaculture implements the photovoltaic power generation control method for power supply in aquaculture when executed by a processor.

[0044] The beneficial effects of this invention are as follows: By introducing a multi-level threshold gating and state inversion joint mechanism based on the energy allocation plan matrix, this invention can flexibly switch power supply paths according to dynamic parameters such as photovoltaic adjustable surplus power, mains voltage stability and battery SOC state, which significantly improves the utilization rate of photovoltaic resources, avoids load power loss problems caused by frequent mains power interruptions or fluctuations, and enhances the stability and adaptability of operation.

[0045] This invention uses a fuzzy correction method based on dynamic feedback of remaining power to perform online feedback correction on the power supply switching command set. This can correct power mismatch problems caused by prediction errors or load fluctuations in real time, improve the accuracy and response efficiency of the switching strategy, and effectively reduce losses caused by battery over-discharge, energy waste and frequent switching. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a schematic flowchart of the first embodiment of a photovoltaic power generation control method for power supply in aquaculture according to the present invention.

[0048] Figure 2 This is a schematic diagram of energy distribution under a predictive enhancement feedforward mechanism, representing a first embodiment of a photovoltaic power generation control method for power supply in aquaculture according to the present invention.

[0049] Figure 3 This is a schematic diagram of a photovoltaic power generation control method for power supply in aquaculture according to the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] Example 1: As Figure 1 The diagram shown is a flowchart illustrating the first embodiment of the photovoltaic power generation control method for power supply in aquaculture according to the present invention.

[0052] In the first embodiment, the photovoltaic power generation control method for power supply in aquaculture includes:

[0053] Step S10: Acquire current status information through the STM32 controller, input the current status information into the preset power supply mode recognition model, and the power supply mode recognition model outputs the current power supply mode status identifier. ;

[0054] It should be noted that in step S10, the current status information includes the current photovoltaic power generation. Battery state of charge Mains voltage Mains frequency and historical illumination trend data The current power supply mode status indicators include photovoltaic power generation priority mode, battery priority mode, grid power supplementation mode, and off-grid power supply mode. The power supply mode recognition model is constructed based on a multi-layer sensing network and a gated recurrent unit network, including an input layer for receiving current state information from the STM32 controller; a feature fusion layer for extracting composite features from the current state information using ReLU of the first layer and LeakyReLU of the second layer; a time-series sensing layer for extracting historical illumination fluctuation features, SOC fallback curve features, and grid power fluctuation cycle features from the composite features using a gated recurrent large network unit; a state mapping layer for mapping the historical illumination fluctuation features, SOC fallback curve features, and grid power fluctuation cycle features from the composite features into a fixed-dimensional state space output and confidence score using a softmax classification function; and an output layer for outputting the current power supply mode status indicator based on the state space output and confidence score. .

[0055] Understandably, by introducing a neural network structure that combines fusion feature extraction with temporal modeling, accurate identification of the current power supply mode can be achieved in power supply environments with highly coupled multi-source state information and strong volatility. Compared to traditional methods that use fixed thresholds or single features (such as voltage or SOC) for pattern segmentation, this recognition model can identify subtle dynamic changes such as abnormal disturbances in mains frequency and short-period pulse fluctuations in light intensity in the temporal dimension, thereby effectively improving the robustness and response sensitivity of pattern recognition.

[0056] For example, data was collected from a farm over 14 consecutive days. Statistics showed that the traditional single-threshold strategy only achieved a 64.5% accuracy rate in identifying the "low mains voltage, medium SOC, and strong photovoltaic fluctuations" scenario as "mains supplement mode," while the accuracy increased to 94.8% after adopting the power supply mode identification model of this invention. In the scenario of "continuous rain, intermittent mains power, and low SOC fluctuations," the traditional solution caused an average power supply switching delay of 3.6 seconds due to misidentification, while the model of this invention had an average response time of 0.92 seconds and a misidentification rate of only 1.7%. Experimental results demonstrate that the identification model of this invention possesses stronger generalization ability and dynamic environmental adaptability, significantly improving the intelligent judgment and safe operation level of photovoltaic systems under complex power grid conditions in aquaculture areas.

[0057] Step S20: Based on the current power supply mode status identifier A state-aware, driven scheduling mechanism is used to execute path selection and load priority mapping tasks, outputting a dynamic energy routing matrix. ;

[0058] It should be noted that the "state-aware driven scheduling mechanism" refers to dynamically executing path selection and load priority allocation based on the photovoltaic power generation, battery energy storage, and grid power supply capabilities corresponding to the power supply mode status identifier, taking into account multiple factors such as the current available energy level, load type, and operating period. The path selection task includes switching decisions such as photovoltaic-load direct power supply path, battery-load indirect power supply path, and grid power supplementation path; the load priority mapping task prioritizes according to historical power supply records, load importance tags, and response sensitivity, constructs a power supply sequence table, and outputs a dynamic energy routing matrix. Each element in the matrix represents the matching relationship between the power supply path and the target load at a specific time and power supply state.

[0059] Understandably, the state-aware scheduling mechanism can intelligently select the optimal power supply path and load target combination based on the actual operating status, thereby avoiding resource waste or load mismatch problems that occur when there are sudden changes in power supply capacity in traditional fixed priority scheduling, and achieving more efficient and stable energy management and resource utilization.

[0060] It should be understood that, compared with traditional power supply scheduling methods based on static rules or fixed priority tables, this step effectively improves the adaptability to complex power supply scenarios (such as rapid fluctuations in sunlight, battery critical state, intermittent mains power outages, etc.) by introducing real-time coupling between state recognition results and scheduling logic, ensuring that important loads can still obtain stable power supply under resource-constrained conditions.

[0061] For example, the experiment simulates a typical rural aquaculture scenario, including water pumps (high power / high priority), aerators (medium power / high priority), lighting (low power / medium priority), and monitoring equipment (low power / high priority). Under strong midday sunlight in summer and with full battery power, the system is identified as "priority mode," and the dynamic energy routing matrix prioritizes the photovoltaic-water pump, photovoltaic-aerator, and photovoltaic-monitoring paths to achieve completely green power supply. However, under the "mains power supplement mode" on cloudy or rainy days when mains power flickers and battery power decreases, the priority of lighting and aerators is automatically reduced, and mains power and battery power are prioritized for water pumps and monitoring equipment.

[0062] Step S30: Based on the dynamic energy routing matrix A predictive-enhanced energy feedforward regulation mechanism is used to implement decision control for photovoltaic power generation, and an energy allocation plan matrix is ​​output.

[0063] It's important to clarify that the "predictive-enhanced energy feedforward regulation mechanism" refers to the following: after acquiring the dynamic energy routing matrix, instead of waiting for deviations in photovoltaic output power or load demand to occur before implementing feedback regulation, it proactively predicts the power supply capacity and load demand for the next period based on historical photovoltaic power generation data, irradiance trends, load operation patterns, and grid power fluctuations. Based on the prediction results, it actively generates the power allocation ratio of photovoltaic, battery, and grid power, forming an energy allocation plan matrix. This mechanism typically includes three core components: First, short-term trend prediction of photovoltaic power generation capacity, considering weather changes, time-based patterns, and the impact of panel temperature rise; second, advance estimation of power demand for various loads, focusing on identifying the power demand of high-priority loads (such as aerator pumps and water pumps) and delayable loads (such as lighting and auxiliary wind turbines); third, adjusting the energy allocation strategy in advance based on the difference between the predicted power supply capacity and total demand, rather than switching or scheduling only after a power shortage occurs.

[0064] It should be understood that, compared to traditional simple feedback control methods that rely on real-time photovoltaic output adjustment or only focus on voltage and current changes, the innovation of this step lies in the addition of a combined predictive judgment and feedforward adjustment mechanism, transforming strategy adjustment from "post-event remediation" to "proactive planning." Traditional methods often trigger mains power supplementation or load shutdown only after insufficient photovoltaic output or battery power drops to a certain threshold, leading to delayed response or even short-term power outages. This invention, however, can generate energy allocation plans several minutes to tens of minutes in advance, avoiding sudden power outages and frequent switching. Furthermore, traditional solutions typically ignore the power supply priority relationships between different loads. This step, through a dynamic energy routing matrix combined with feedforward adjustment, ensures that photovoltaic power supply prioritizes critical equipment such as oxygenation and water pumps, intelligently reducing delayed loads and effectively improving overall energy utilization efficiency.

[0065] For example, such as Figure 2 As shown, each small square represents the energy allocation strategy for a certain device in a certain time slice; green indicates photovoltaic priority power supply (predictive enhancement feedforward target); yellow indicates battery regulation (feedforward compensation); red indicates mains power backup (feedforward prediction avoids frequent use); the figure clearly shows that the feedforward mechanism preemptively adjusts battery and mains power for allocation when photovoltaic power is insufficient, thereby ensuring that high-priority loads do not run uninterrupted.

[0066] Step S40: Based on the energy allocation plan matrix, a multi-level threshold gating and state inversion joint mechanism is used to perform power supply switching control processing, and output the power supply switching instruction set;

[0067] It should be noted that the "multi-level threshold gating and state inversion joint mechanism" refers to a decision-making mode that, during the power supply switching control process, does not rely on a single voltage, current, or power value to trigger switching, but introduces a hierarchical dynamic threshold judgment system and a collaborative decision-making mode of operating state inversion logic. Specifically, this mechanism mainly includes the following three aspects: First, under different types of loads and different operating periods, multiple levels of threshold parameters such as voltage, current, and SOC (battery state of charge) are preset to finely classify power supply conditions; Second, when the threshold triggering critical boundary is near, the power supply switching is not immediately executed, but it is judged whether the current state is a trend of historical state inversion (e.g., from charging to discharging, from power increase to decrease), thereby enhancing the timing robustness of the judgment; Third, when both the threshold judgment and state inversion are satisfied, a switching command is generated, thereby effectively avoiding frequent switching caused by short-term fluctuations, peak interference, or false alarms.

[0068] Understandably, by introducing multi-level threshold gating and state inversion for joint judgment, this step enhances the fault tolerance and decision stability of the power supply switching strategy. For example, when the photovoltaic output is near a critical value but still on an upward trend, photovoltaic power supply will continue to be maintained instead of immediately switching to battery or mains power. This trend-based judgment method reduces the number of switching cycles in a short period of time, avoiding the overhead caused by electrical component wear and frequent execution of control logic.

[0069] Step S50: Apply the power supply switching instruction set and use the fuzzy correction method based on the dynamic feedback of the remaining power to perform feedback control of photovoltaic power generation, and output the optimized power supply switching instruction set.

[0070] It should be noted that the "fuzzy correction method based on dynamic feedback of remaining power" refers to the real-time collection of operating indicators such as the remaining power of photovoltaic systems, grid connection status, battery SOC, and operating current and frequency of motor loads in aquaculture photovoltaic systems, based on the currently executed power supply switching command set. This data is then used to perform corrective adjustments through a fuzzy controller. The controller's rule base is customized for the characteristics of aquaculture loads, particularly considering the power supply sensitivity of critical loads requiring continuous operation, such as aerators and water pumps, and auxiliary loads such as lighting and fans that can tolerate short-term power reductions. It uses "remaining power amplitude," "load priority," and "grid stability" as primary input factors, combining fuzzy rules to infer whether to adjust the upcoming switching action or maintain the current power supply status, thus forming a flexible and adjustable control correction mechanism.

[0071] It is understandable that introducing dynamic residual power sensing capabilities and using fuzzy logic can help avoid premature or frequent switching. In aquaculture scenarios, fluctuations in sunlight and load changes are often influenced by both weather and operational activities. Traditional control strategies based on fixed thresholds are prone to triggering frequent power switching due to "critical fluctuations," leading to unstable power supply or equipment wear. However, by adopting the fuzzy correction method in this step, a more reasonable correction judgment can be made based on the current "power redundancy" of the photovoltaic system to the load. For example, when the photovoltaic power is slightly below the threshold but the battery is sufficiently charged, the mains power switch can be delayed; or when there is ample power redundancy, photovoltaic power can be prioritized to ensure critical loads while reducing auxiliary load power, thereby effectively reducing unnecessary switching actions and improving photovoltaic energy utilization and operational stability.

[0072] For example, consider a large aquaculture farm whose load structure includes 4 aerators (operating day and night), 2 water pumps (operating on a timer), 6 sets of LED lighting (operating at night), and 1 auxiliary fan. During a typical daily cycle of cloudy to sunny weather, the traditional control system triggered 5 mains power switching actions between 8:45 AM and 9:30 AM due to short-term fluctuations in sunlight, causing severe relay jitter and intermittent power meter input. By employing the fuzzy correction mechanism of this invention, it automatically identifies that although the photovoltaic power is in a downward trend, the State of Charge (SOC) remains stable, and predicts that sunlight will stabilize after 9:00 AM. Therefore, the switching is delayed, and photovoltaic output finally resumes after 9:15 AM, eliminating the need for mains power connection. The proportion of photovoltaic power supply throughout the day increases by approximately 18.7%, the number of power supply switching times decreases to less than 2, the continuity of load operation is significantly enhanced, and electricity costs are saved by approximately 9.3%.

[0073] Example 2: Furthermore, the present invention provides a photovoltaic power generation control system for aquaculture, employing a photovoltaic power generation control method for aquaculture as described in the above embodiments, which can solve a technical problem related to photovoltaic power generation control for aquaculture. Compared with the prior art, the beneficial effects of the photovoltaic power generation control system for aquaculture provided by the present invention are the same as those of the photovoltaic power generation control method for aquaculture provided in the above embodiments, and other technical features of the photovoltaic power generation control system for aquaculture are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0074] Example 3: This invention provides a photovoltaic power generation control device for power supply in aquaculture. Please refer to... Figure 3A photovoltaic power generation control device for aquaculture includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to execute the photovoltaic power generation control method for aquaculture described in Embodiment 1 above. The photovoltaic power generation control device for aquaculture in this embodiment may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), vehicle terminals (e.g., vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. This photovoltaic power generation control device for aquaculture is merely an example and should not limit the functionality or scope of the embodiments of this invention. The photovoltaic power generation control device for aquaculture may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory 1002 or a program loaded from a storage device 1003 into a random access memory 1004. Random access memory 1004 also stores various programs and data required for the operation of a photovoltaic power generation control device for aquaculture. Processing device 1001, read-only memory 1002, and random access memory 1004 are interconnected via bus 1005. I / O interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows a photovoltaic power generation control device for aquaculture to communicate wirelessly or wiredly with other devices to exchange data. Although a photovoltaic power generation control device for aquaculture with various systems is shown in the figure, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented or possessed alternatively.

[0075] Example 4: This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the photovoltaic power generation control method for power supply in aquaculture as described above. The computer program product provided by this invention can solve a technical problem related to photovoltaic power generation control for power supply in aquaculture. Compared with the prior art, the beneficial effects of the computer program product provided by this invention are the same as those of the photovoltaic power generation control method for power supply in aquaculture provided in the above embodiments, and will not be repeated here.

[0076] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this invention.

[0077] It should be understood that the various parts disclosed in this invention can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0078] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A photovoltaic power generation control method for power supply for aquaculture, characterized by, The method comprises: Step S10: Collect the current state information through the STM32 controller, input the current state information into a preset power supply mode recognition model, and output a current power supply mode state identifier by the power supply mode recognition model ; Step S20: identifying the current power supply mode state based on the power supply mode state The path selection task and the load priority mapping task are executed by using a state-aware driving scheduling mechanism, and a dynamic energy routing matrix is output ; Step S30: According to the dynamic energy routing matrix The decision control of photovoltaic power generation is performed by using a prediction-enhanced energy feedforward adjustment mechanism, and an energy distribution plan matrix is output. Step S40: based on the energy distribution plan matrix, a multi-level threshold gating and state inversion combined mechanism is used for power supply switching control processing, and a power supply switching instruction set is output; Step S50: applying the power supply switching instruction set, a fuzzy correction method based on residual power dynamic feedback is used to execute feedback control of photovoltaic power generation, and an optimized power supply switching instruction set is output.

2. The photovoltaic power generation control method for power supply for farming according to claim 1, characterized by, In step S10, the current state information includes current photovoltaic power , battery state of charge , mains voltage , mains frequency and historical illumination trend data ; the current power supply mode state identifier includes a photovoltaic power generation priority mode, a battery priority mode, a mains power supplement mode and an off-grid power supply mode; the power supply mode recognition model is constructed based on a multi-layer perception network and a gated recurrent unit network, and includes an input layer for receiving the current state information from the STM32 controller; The feature fusion layer is configured to extract composite features in the current state information by using first layer full connection neurons ReLU and second layer full connection neurons LeakyReLU; the time sequence perception layer is configured to extract historical light fluctuation features, SOC falling curve features and city power fluctuation period features in the composite features by using gated recurrent unit cells; the state mapping layer is configured to map the historical light fluctuation features, the SOC falling curve features and the city power fluctuation period features in the composite features into a fixed-dimension state space output and a confidence by using a softmax classification function; and the output layer is configured to output a current power supply mode state identifier according to the state space output and the confidence. .

3. The photovoltaic power generation control method for power supply for aquaculture according to claim 2, characterized by, In step S20, based on the current power supply mode state identifier The state-aware driving scheduling mechanism is adopted to perform the path selection task and the load priority mapping task, and output a dynamic energy routing matrix The step specifically comprises: Step S201: First, the current power supply mode state is identified based on Perform path selection task: If the current power supply mode state identifier is the photovoltaic power generation priority mode, the first path is activated, and the first path is represented as ; If the current power supply mode state identifier is the battery priority mode, then the second path is activated, the second path is represented as ; If the current power supply mode state identifier is the mains supply supplement mode, the third path is activated, and the third path is represented as ; If the current power supply mode state identifier is the off-grid power supply mode, the fourth path is activated, and the fourth path is represented as ; wherein, represents a photovoltaic power source; represents a battery discharge path; represents a utility power source; represents a high priority load; represents a battery charging path; represents a low priority load; represents direct supply of a primary priority load by the photovoltaic power source; represents priority use of the photovoltaic power source for battery charging; represents supply of a primary priority load by battery discharge; represents supply of a low priority load by battery discharge; represents direct supply of a primary priority load by utility power; represents direct supply of a low priority load by utility power; represents independent supply of a critical load by the battery under off-grid conditions; Step S202: The load priority mapping task is performed again: the current load power is obtained, the path activation priority rule is constructed based on the current load power; the path activation priority rule is used for dynamic path priority mapping processing for the first path, the second path, the third path and the fourth path, and finally a dynamic energy routing matrix is output .

4. The photovoltaic power generation control method for power supply for farming according to claim 1, characterized by, In step S30, according to the dynamic energy routing matrix The step of outputting the energy distribution plan matrix by using the prediction-enhanced energy feedforward adjustment mechanism to perform the decision control of photovoltaic power generation specifically includes: Step S301: obtaining the path of the photovoltaic power supply from the dynamic energy routing matrix The path of the photovoltaic power supply is obtained from the dynamic energy routing matrix, and the photovoltaic power supply path is constructed; the current photovoltaic power generation power in the photovoltaic power supply path is collected, and if all the current photovoltaic power generation powers are greater than the preset active load power requirement, the photovoltaic power supply path is marked as a selectable energy path. Step S302: based on the optional energy path, a segmented weighted trend fitting method is used for photovoltaic power generation prediction calculation, and the next moment photovoltaic prediction power is output; Step S303: when the photovoltaic prediction power is less than the preset photovoltaic power generation threshold, the battery charging current of the battery charging path in the optional energy path is set to the preset maximum charging current; when the photovoltaic prediction power is greater than or equal to the preset photovoltaic power generation threshold, the battery charging current of the battery charging path in the optional energy path is set to 0.5 times the preset maximum charging current; finally, the energy distribution plan matrix is output, which is used to indicate the power distribution of photovoltaic power supply, battery power supply and mains power supply to each load path.

5. The photovoltaic power generation control method for power supply for aquaculture according to claim 4, characterized in that, In step S30, based on the optional energy path, a segmented weighted trend fitting method is used for photovoltaic power generation prediction calculation, and the next moment photovoltaic prediction power is output, which specifically comprises: Obtain historical photovoltaic prediction power data, and extract short-period photovoltaic power generation change characteristics based on the short sliding window gradient extraction principle of the historical photovoltaic prediction power data; extract long-period photovoltaic power generation change characteristics based on the long sliding window linear fitting principle of the historical photovoltaic prediction power data; construct short-period change rate parameters and long-period trend slope parameters based on the short-period photovoltaic power generation change characteristics and the long-period photovoltaic power generation change characteristics, respectively; Based on the short-period change rate parameters and the long-period trend slope parameters, a weighted nonlinear combination fitting method is used to construct a weighted prediction function, and the weighted prediction function outputs a preliminary photovoltaic prediction power; The preliminary photovoltaic prediction power is subjected to error suppression processing by using the offset confidence constraint method, and the final next moment photovoltaic prediction power is output.

6. The photovoltaic power generation control method for supplying power for aquaculture according to claim 1, wherein In step S40, based on the energy distribution plan matrix, a multi-level threshold gating and state inversion combined mechanism is used for power supply switching control processing, and a power supply switching instruction set is output, which specifically comprises: Step S401: Extract the target load set and the current resource state from the energy distribution plan matrix, and introduce a multi-level dynamic judgment threshold, which includes a photovoltaic available power upper threshold , a photovoltaic available power lower threshold , a city power voltage stability upper threshold , a city power voltage stability lower threshold , and a battery SOC standby threshold ; Based on the multi-level dynamic judgment threshold, the power supply path enable state in the energy distribution plan matrix is judged in real time, and the path candidate state set with the power supply path enable state being normal is output. Step S402: based on the path candidate state set, a steady-state probability evaluation mechanism based on a Markov state transition graph is used to execute path switching decision judgment, and a power supply path confirmation set is output; Step S403: according to the power supply path confirmation set, a preset finite state machine FSM is used to dynamically generate a power supply switching instruction set, which is used to control the circuit breaker and the relay to execute path start-stop operation; the state of the finite state machine FSM is defined as photovoltaic power supply state, mains power supply state, battery power supply state and hybrid power supply state.

7. The photovoltaic power generation control method for supplying power for aquaculture according to claim 1, wherein In step S50, the power supply switching instruction set is applied, and a fuzzy correction method based on residual power dynamic feedback is used to execute feedback control of photovoltaic power generation, and an optimized power supply switching instruction set is output, which specifically comprises: Step S501: Collect the actual power consumption of the load and the actual output power of the photovoltaic after the application of the power supply switching instruction set, and calculate the photovoltaic remaining adjustable power based on the actual power consumption of the load and the actual output power of the photovoltaic ; if , it indicates that the photovoltaic power supply is insufficient, and the mains or battery needs to be supplemented; if , it indicates that there is still photovoltaic redundancy; Step S502: based on the photovoltaic remaining adjustable power The dynamic fluctuation adjustment factor is constructed, and a sensitivity correction operation of the power supply path switching strategy is performed based on the dynamic fluctuation adjustment factor by using the fuzzy rule controller to output a control adjustment factor set; wherein the control adjustment factor set is used to adjust the trigger boundary and the waiting time threshold value of the path switching. Step S503: Perform fuzzy correction operation on the power supply switching instruction set according to the control adjustment factor set, including dynamic delay trigger, path priority adjustment and boundary gate value correction processing, and finally output the optimized power supply switching instruction set.

8. A photovoltaic power generation control system for power supply for aquaculture, applied to the photovoltaic power generation control method for power supply for aquaculture in any one of claims 1 to 7, characterized in that, The photovoltaic power generation control system for power supply in aquaculture comprises: The power supply state recognition module is configured to collect current state information through an STM32 controller, input the current state information into a preset power supply mode recognition model, and output a current power supply mode state identifier by the power supply mode recognition model ; a path and priority scheduling module configured to identify a current power supply mode state based on The path selection task and the load priority mapping task are executed by using a state-aware driving scheduling mechanism, and a dynamic energy routing matrix is output ; An energy feedforward regulation module is configured to generate a dynamic energy routing matrix based on the energy demand matrix and the energy supply matrix The decision control of photovoltaic power generation is performed by using a prediction-enhanced energy feedforward regulation mechanism, and an energy distribution plan matrix is output. The power supply switching control module is configured to perform power supply switching control processing based on the energy distribution plan matrix by using a multi-level threshold gate and state inversion joint mechanism, and output a power supply switching instruction set. The feedback correction control module is configured to apply the power supply switching instruction set and perform feedback control of photovoltaic power generation by using a fuzzy correction method based on residual power dynamic feedback, and output an optimized power supply switching instruction set.

9. A photovoltaic power generation control device for power supply for aquaculture, characterized by, The photovoltaic power generation control device for power supply in aquaculture comprises a memory, a processor, and a photovoltaic power generation control program for power supply in aquaculture stored in the memory and executable on the processor. When the photovoltaic power generation control program for power supply in aquaculture is executed by the processor, the photovoltaic power generation control method for power supply in aquaculture described in any one of claims 1 to 7 is implemented.

10. A computer program product, characterised in that, The computer program product comprises a photovoltaic power generation control program for power supply in aquaculture. When the photovoltaic power generation control program for power supply in aquaculture is executed by a processor, the photovoltaic power generation control method for power supply in aquaculture described in any one of claims 1 to 7 is implemented.

Images