Photovoltaic parking building power supply management method, device and system, and storage medium

Abstract

The invention relates to the technical field of photovoltaic power supply management, and discloses a photovoltaic parking building power supply management method, device and system, and a storage medium. The method comprises the steps of determining real-time irradiation and component temperature data of a parking building ceiling photovoltaic component matrix, calculating an actual light receiving area according to component surface shadow distribution, and generating a photovoltaic output power correction parameter in combination with a historical attenuation coefficient; constructing a photovoltaic power generation capacity fluctuation sequence in unit time based on the parameters, identifying a power generation capacity decline time period and a power output gap, and marking the power generation capacity decline time period and the power output gap as a power supply unstable interval; according to the interval time sequence distribution, matching real-time load demand data of the charging piles in the building, calculating the power difference between the load demand and the power generation capacity, and generating a priority power supply sequence table; and dynamically distributing photovoltaic power to charging loads with different priorities according to the sequence table, adjusting charging and discharging threshold values of the building energy storage system, and generating a photovoltaic power dispatching instruction. The method can accurately evaluate the power generation capacity, respond to power generation fluctuation and optimize power distribution.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photovoltaic power supply management, in particular to a photovoltaic parking building power supply management method, device and system, and a storage medium. BACKGROUND

[0002] With the rapid development of new energy industry, photovoltaic technology is increasingly widely used in the field of buildings. As a new type of building form integrating parking function and photovoltaic power generation function, photovoltaic parking buildings have gradually become an important part of urban energy systems. Such buildings can convert solar energy into electricity by arranging photovoltaic modules in a square array on the roof, providing power support for charging piles and other loads in the building, and playing an important role in relieving urban power pressure and promoting clean energy consumption.

[0003] However, there are still many problems to be solved in the current power supply management process of photovoltaic parking buildings. In the photovoltaic power generation link, the output power of photovoltaic modules is easily affected by external environmental factors. The fluctuation of real-time irradiance, the change of module temperature, and the shadow distribution formed by the surrounding obstacles will all cause the actual light receiving area of the module to change, resulting in unstable output power. At the same time, photovoltaic modules will experience performance degradation during long-term use. If the historical degradation coefficient is not fully considered, it will be difficult to accurately assess the actual power generation capacity, making it difficult to schedule power supply.

[0004] In the power supply scheduling and load matching link, the traditional management method mostly adopts a fixed power supply mode, lacking real-time response to the dynamic changes of photovoltaic power generation capacity. When the photovoltaic power generation capacity decreases, forming a power output gap, it is difficult to identify the unstable power supply interval in time, and it is also difficult to reasonably allocate limited photovoltaic power according to the real-time load demand difference of charging piles. This situation not only may cause some charging piles to be unable to work normally due to insufficient power supply, affecting user experience, but also may cause waste of photovoltaic power or excessive dependence on grid power supply, which is not conducive to maximizing energy utilization efficiency. In addition, the charge and discharge threshold of the building energy storage system is usually fixed and cannot be dynamically adjusted according to the real-time changes of photovoltaic power generation capacity and load demand, further reducing the flexibility and economy of the entire power supply system, and restricting the stable and efficient operation of the photovoltaic parking building power supply system. SUMMARY

[0005] The present application aims to provide a photovoltaic parking building power supply management method, device, system and storage medium to solve the problems raised in the background.

[0006] To achieve the above-mentioned purpose, the present application provides a photovoltaic parking building power supply management method, which comprises: Determine real-time irradiation data and component temperature data of the photovoltaic module square array on the parking building roof, calculate the actual light receiving area according to the shadow distribution state of the component surface, and generate a photovoltaic output power correction parameter in combination with the historical attenuation coefficient; Based on the photovoltaic output power correction parameter, construct a photovoltaic power generation capacity fluctuation sequence in a unit time, identify the power generation capacity decline period and power output gap, and mark it as a power supply unstable interval; According to the time sequence distribution corresponding to the power supply unstable interval, match the real-time load demand data of the charging pile in the parking building, calculate the power difference between the load demand and the power generation capacity, and generate a priority power supply sequence table; According to the sorting result in the priority power supply sequence table, dynamically allocate photovoltaic power to different priority charging loads, adjust the charge and discharge threshold of the building energy storage system, and generate a photovoltaic power dispatching instruction.

[0007] Preferably, the photovoltaic output power correction parameter includes the effective light receiving proportion of the component, the temperature loss coefficient, and the shadow shielding influence factor, the power supply unstable interval includes the output power decline duration, the power fluctuation amplitude, and the expected power shortage, the priority power supply sequence table includes the load emergency degree identifier, the power allocation weight coefficient, and the energy storage compensation start condition, and the photovoltaic power dispatching instruction includes the energy storage release time window, the load power supply interruption protection flag, and the power dynamic balance range.

[0008] Preferably, the determination step of the photovoltaic output power correction parameter is specifically: Collect the irradiation intensity values and component backboard temperature values of the photovoltaic module square array on the parking building roof at multiple sampling times, scan the shadow distribution form of the component surface, calculate the area proportion of the shadow shielding area and the shielding duration, and calculate the actual output power loss proportion in combination with the component nominal power and the temperature characteristic curve according to the area proportion of the shadow shielding area and the shielding duration. According to the area proportion of the shadow shielding area and the shielding duration, calculate the actual output power loss proportion in combination with the component nominal power and the temperature characteristic curve, and generate an irradiation influence feature set; According to the power loss proportion in the irradiation influence feature set, introduce the historical running attenuation coefficient of the component and the current environmental cleanliness index, comprehensively calculate the real-time output capacity correction value of the photovoltaic component, and generate the photovoltaic output power correction parameter.

[0009] Preferably, the construction step of the power supply unstable interval is specifically: Call the real-time output capacity correction value in the photovoltaic output power correction parameter, continuously record the fluctuation of the photovoltaic power generation power in multiple periods, identify the duration and power drop amplitude when the power generation power is lower than a preset threshold, and judge the severity of the photovoltaic power supply shortage in combination with the current state of charge and the adjustable capacity of the building energy storage system. Statistically summarize the duration and power drop amplitude when the power generation power is lower than a preset threshold, and judge the severity of the photovoltaic power supply shortage in combination with the current state of charge and the adjustable capacity of the building energy storage system. Based on the severity of the photovoltaic power supply shortage, the starting point and expected duration of the power supply shortage are marked, and the expected load increase of the charging piles during this period are associated to generate a power supply instability interval.

[0010] Preferably, the steps for generating the priority power supply order table are as follows: Obtain the timing information corresponding to the unstable power supply interval, synchronously collect the real-time load power and vehicle battery charging status data of each charging pile in the parking building, and calculate the remaining charging energy and time sensitivity of each charging pile. Based on the remaining charging energy required for each charging station and time sensitivity, combined with the vehicle's scheduled departure time and the battery's safe charging threshold, the urgency and interruptibility of the charging task are assessed. Based on the urgency and interruptibility of the charging task, a power supply priority number and power allocation weight are assigned to each charging load to generate a priority power supply order table.

[0011] Preferably, the generation step of the photovoltaic power dispatch instruction specifically includes: Based on the power supply priority number and power allocation weight in the priority power supply order table, the allocation ratio of photovoltaic power to high-priority charging loads is determined, and the discharge power and duration of the building energy storage system are adjusted at the same time. Monitor the discharge power and duration of the building energy storage system, and dynamically adjust the power supply parameters of different priority loads based on real-time photovoltaic power generation fluctuations. Based on the dynamically adjusted power supply parameters of loads with different priorities, a photovoltaic power dispatch instruction containing power allocation instructions and energy storage control strategies is generated.

[0012] Preferably, the method further includes the following steps: After executing the photovoltaic power dispatch command, continuously monitor the actual power supply stability and battery charging efficiency of high-priority charging loads, and record power interruption events and power deviation values; Analyze the distribution patterns of the power outage events and power deviation values, and revise the prediction model for subsequent unstable power supply intervals and the generation logic of the priority power supply order table. Based on the revised prediction model and generation logic, the power dynamic balance range and load protection conditions in the photovoltaic power dispatch instructions are updated.

[0013] Preferably, the present invention further includes a photovoltaic parking building power supply management device for implementing the photovoltaic parking building power supply management method described above, the system comprising: The data acquisition module is used to acquire real-time irradiance data and module temperature data of the photovoltaic module array on the roof of the parking building; The shadow distribution calculation module is used to calculate the actual light-receiving area based on the shadow distribution state of the component surface; The power correction parameter generation module is used to generate photovoltaic output power correction parameters by combining historical attenuation coefficients. A power generation capacity fluctuation sequence construction module is used to construct a photovoltaic power generation capacity fluctuation sequence per unit time based on the photovoltaic output power correction parameters. The power supply instability interval identification module is used to identify periods of reduced power generation capacity and power output gaps and mark them as power supply instability intervals; The load demand matching module is used to match the real-time load demand data of charging piles in the parking building according to the power supply instability range. The power difference calculation module is used to calculate the power difference between load demand and power generation capacity. A priority power supply order table generation module is used to generate a priority power supply order table based on the power difference; The power dispatch module is used to dynamically allocate photovoltaic power to charging loads of different priorities according to the priority power supply order table, adjust the charging and discharging thresholds of the building energy storage system, and generate photovoltaic power dispatch instructions.

[0014] Preferably, the present invention also includes a photovoltaic parking building power supply management system, the system including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor, when executing the computer program, implements the steps of the photovoltaic parking building power supply management method described above.

[0015] Preferably, the present invention further includes a storage medium storing a computer program thereon, which, when executed by a processor, implements the photovoltaic parking building power supply management method as described above.

[0016] Compared with the prior art, the beneficial effects of the present invention are: In terms of photovoltaic power generation assessment, by comprehensively collecting real-time irradiance data and module temperature data of photovoltaic module arrays on the roof of parking buildings, combining the actual light-receiving area with the shading distribution of the module surface, and incorporating historical attenuation coefficients to generate photovoltaic output power correction parameters, the actual power generation capacity of photovoltaic modules can be more accurately reflected, reducing the impact of external environmental factors and module performance degradation on power generation assessment results, and providing reliable basic data support for subsequent power supply scheduling.

[0017] In terms of identifying and responding to unstable power supply intervals, a photovoltaic power generation capacity fluctuation sequence is constructed based on photovoltaic output power correction parameters within a unit time period. This allows for timely and accurate identification of periods of declining power generation capacity and power output gaps, thus clearly defining unstable power supply intervals. This dynamic identification method breaks through the limitations of traditional fixed power supply models, enabling power supply management to proactively adapt to changes in photovoltaic power generation capacity, prepare for power allocation in advance, avoid power supply chaos caused by sudden drops in power generation capacity, and ensure the overall stability of the power supply system.

[0018] In the load matching and power distribution stage, based on the temporal distribution of unstable power supply intervals, the real-time load demand data of charging piles is matched and the power difference is calculated to generate a priority power supply order table, enabling differentiated allocation of photovoltaic power. Allocating power to charging loads that need it most according to priority ensures that critical or high-demand charging piles receive priority power support, reducing charging pile outages due to insufficient power, improving user experience, and avoiding power waste caused by indiscriminate power supply. This maximizes the utilization of limited photovoltaic power, reduces dependence on grid power, and saves energy costs.

[0019] Regarding the coordinated operation of energy storage systems, the charging and discharging thresholds of building energy storage systems are dynamically adjusted according to the priority power supply order, generating photovoltaic power dispatch instructions to fully leverage the regulatory role of energy storage systems. When photovoltaic power generation capacity is sufficient and load demand is low, the excess photovoltaic power is stored by increasing the charging threshold of the energy storage system; when photovoltaic power generation capacity is insufficient and a power gap occurs, the discharging threshold is lowered to release energy storage power to supplement the power supply, achieving a dynamic balance between photovoltaic power generation, energy storage, and load demand. This flexible charging and discharging adjustment method not only improves the utilization efficiency of energy storage systems but also further enhances the flexibility and economy of the power supply system, reduces energy loss, and promotes the operation of photovoltaic parking building power supply systems towards a more efficient, stable, and economical direction. Attached Figure Description

[0020] Figure 1 This is a schematic diagram illustrating the working principle of the photovoltaic parking building power supply management method described in this invention. Figure 2 Flowchart for determining correction parameters for photovoltaic output power; Figure 3 A flowchart for generating a priority power supply order table. Detailed Implementation

[0021] 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.

[0022] Please see Figure 1 This invention provides a method for power supply management of photovoltaic parking buildings. The method includes: acquiring real-time irradiance data and backsheet temperature data of the photovoltaic module array on the roof of the parking building; monitoring the shadow distribution pattern on the module surface using an image sensor or laser scanning device; calculating the actual light-receiving area ratio; and generating correction parameters reflecting the current power generation capacity by combining the attenuation coefficient from the historical operating records of the photovoltaic modules. This step relies on environmental monitoring equipment and a data processing unit. Real-time irradiance data is collected by a radiometer, module temperature data is obtained by a temperature sensor, shadow distribution is captured by a camera or scanner, the actual light-receiving area is calculated using an image processing algorithm, the historical attenuation coefficient is retrieved from a database, and the correction parameters are calculated by a microprocessor.

[0023] Based on the generated photovoltaic output power correction parameters, the system constructs a photovoltaic power generation capacity fluctuation sequence per unit time. The sampling interval can be set to the second or minute level. It continuously records power values ​​and analyzes their changing trends, identifying periods where power generation capacity is lower than a preset threshold. These periods may be due to cloud cover or shadow movement causing power output gaps and are marked as unstable power supply intervals. The identification of unstable power supply intervals involves time series analysis algorithms, such as sliding window averaging or anomaly detection models. The threshold is dynamically adjusted based on the rated power of the photovoltaic modules and the building's load demand. Subsequently, the system matches the real-time load demand data of charging piles within the parking building with the start and end times of the unstable power supply intervals. Load data, collected through smart meters or charging pile controllers, includes charging power, vehicle battery charge, and estimated charging time. The system calculates the power difference between power generation capacity and load demand, obtained through subtraction. Based on the difference size and load characteristics, a priority power supply order table is generated. The priority power supply order table sorts the charging loads based on factors such as urgency and interruptibility.

[0024] The system dynamically allocates photovoltaic (PV) power to charging loads of different priorities based on the priority power supply order table. High-priority loads are supplied first, while low-priority loads may be restricted or interrupted. Simultaneously, it adjusts the charging and discharging thresholds of the building's energy storage system, such as setting the upper limit of the storage's discharge power and duration, and generates PV power dispatch instructions containing specific control commands. These dispatch instructions are sent to charging piles and energy storage systems via communication modules to achieve real-time power balance. Throughout the process, data flow is integrated through an IoT platform to ensure real-time performance and accuracy. The update frequency of PV output power correction parameters can be adjusted according to environmental changes. The marking of unstable power supply intervals considers historical data patterns. Machine learning algorithms are incorporated into the generation of the priority power supply order table to optimize the sorting logic. The execution of power dispatch instructions is coordinated by a central controller, reducing dependence on the power grid.

[0025] Example 1: See Figure 2 The determination of photovoltaic output power correction parameters is based on the synchronous acquisition and fusion calculation of multi-source data. This process begins with a comprehensive perception of the operating status of the photovoltaic module array on the roof of the parking building. A radiometer deployed at a specific location on the roof continuously measures the irradiance reaching the module surface at a set sampling frequency. This frequency can be dynamically adjusted according to weather conditions; for example, the sampling rate is automatically increased during the afternoon when sunlight changes drastically. Working in tandem, a temperature sensor array is securely installed on the back panel of representative modules to accurately capture their operating temperature. These physical measurement data are transmitted in real time to the central processing unit via a fieldbus network. To quantify the shading effect, a visual monitoring system is activated. This system consists of multiple high-definition cameras, whose placement is optimized to cover the entire module array. The cameras continuously acquire image data from the module surface, and image recognition algorithms analyze the distribution pattern of shadows. The algorithm can distinguish between instantaneous shadows caused by moving clouds and fixed shadows cast by nearby buildings or trees, accurately outlining the contours of the shadowed areas. The calculation of the actual light-receiving area is not simply the total area of ​​the module minus the shadow area. Instead, it is necessary to consider the transmission characteristics of the shadow and its impact on the series and parallel circuits of the module. Therefore, the calculation process introduces the arrangement topology of the cells inside the module, maps the shadow coverage area to the specific cell string, and then obtains the effective light-receiving area ratio. This ratio is a dynamically changing parameter.

[0026] Based on the calculated area proportion of the shaded region and its duration, the system needs to further assess its actual impact on power generation capacity. This assessment closely integrates the nominal power parameters of the photovoltaic module and its inherent temperature characteristic curve. The nominal power is taken from the module's manufacturer's technical specifications, while the temperature characteristic curve describes the quantitative relationship between output power and backsheet temperature increase, typically expressed as a temperature coefficient. Calculating the actual output power loss ratio is a multi-factor coupled analysis process. The algorithm comprehensively considers the combined effects of current irradiance, module temperature, and shading. For example, under high temperature and strong light conditions, the same shading may cause more significant power loss. The calculation model generates an irradiance impact feature set containing a time series, recording the power degradation caused by environmental factors at each sampling moment. To improve the accuracy of the correction parameters, the performance degradation caused by long-term module operation must be taken into account. The historical operating degradation coefficient is obtained by analyzing the deviation between accumulated power generation data and initial nominal data, reflecting the slow efficiency decline trend caused by factors such as module material aging and potential microcracks.

[0027] The current environmental cleanliness index is another important correction factor. Contaminants such as dust, pollen, and bird droppings on the module surface reduce light transmittance. This index can be directly measured by an optical dust sensor installed on-site, or indirectly estimated by analyzing the long-term deviation between theoretical and actual power generation under ideal, completely clean, and shadow-free conditions. The comprehensive calculation of the real-time output capacity correction value of the photovoltaic module is a data fusion process. The processing unit weights and fuses the irradiance impact feature set, historical attenuation coefficient, and cleanliness index. The weighting coefficients can be dynamically allocated according to the sensitivity of each factor to the output power. The final photovoltaic output power correction parameter is a comprehensive set of indicators. This parameter set explicitly includes key sub-parameters such as the module's effective light-receiving ratio, temperature loss coefficient, and shading impact factor. The module's effective light-receiving ratio is directly derived from previous image analysis results. The temperature loss coefficient is obtained by querying the temperature characteristic curve value corresponding to the current backsheet temperature. The shading impact factor is a quantitative parameter that integrates the shading area, duration, and position on the module circuitry. These sub-parameters are encapsulated in a data structure, providing accurate input for subsequently constructing a power generation capacity fluctuation sequence.

[0028] The entire parameter determination process forms a closed-loop system. New monitoring data is continuously fed back to fine-tune the calculation model and coefficients, ensuring that the corrected parameters continuously track the actual operating status of the photovoltaic module array. Data acquisition synchronization is crucial. The system clock ensures strict alignment of the three sets of data—irradiance, temperature, and image—in terms of timestamps, avoiding calculation errors caused by time asynchrony. Data transmission uses an industrial communication protocol with a verification mechanism, guaranteeing data integrity and reliability. The calculation process is deployed on edge computing gateways or local servers with sufficient computing power to meet real-time requirements. For large parking complexes, calculation tasks can be distributed across multiple nodes, improving the overall system processing efficiency. After parameter generation, they are accurately timestamped and stored in a time-series database for subsequent modules to query and access by time interval. The storage of historical data also provides a data foundation for long-term performance analysis and model optimization. The shading distribution calculation module and the power correction parameter generation module exchange data through a clearly defined application programming interface, ensuring decoupling and maintainability between system modules.

[0029] Example 2: The construction process of the unstable power supply interval begins with the invocation of photovoltaic output power correction parameters. These parameters contain the component output capacity assessment value after real-time environmental factor correction. The system continuously reads this correction value at a fixed time step and records its changes according to a set period length, thus forming a photovoltaic power generation capacity fluctuation sequence. The length of each period can be configured according to the needs of the management strategy, such as using a 15-minute or 1-hour rolling time window. During the construction of the fluctuation sequence, the system continuously compares the real-time power generation capacity with a dynamically set preset threshold. This threshold is not fixed but is a floating threshold value determined by comprehensively considering the photovoltaic system's rated power, historical average output, and the current overall building electricity load level. When the data points in the fluctuation sequence are below this threshold for multiple consecutive periods, the identification of abnormal periods is triggered. The identification work not only focuses on the simple fact that the power generation is below the threshold but also analyzes the duration of this state and the specific magnitude of the power decline within this period. The duration begins when the power first falls below the threshold and continues until it rebounds and stabilizes above the threshold. The magnitude of the power decline is quantified by calculating the difference between the peak and trough values ​​of the power curve within this duration.

[0030] Statistical analysis of the duration of power shortages below preset thresholds and their corresponding power reductions forms the basis for assessing system operational status. However, isolated power data is insufficient to accurately determine the severity of the power supply gap. Therefore, real-time status information from building-integrated energy storage systems must be incorporated as a key decision-making basis. The current state of charge of the energy storage system reflects its readily available energy reserves, while dispatchable capacity indicates the maximum supporting power that the energy storage converter can provide in a short period. Combining the photovoltaic power shortage with the supporting potential of the energy storage system, the system performs a comprehensive assessment of the severity of the photovoltaic power supply shortage. The assessment logic is not a simple binary classification but employs a multi-level evaluation system. For example, severity is divided into different levels such as mild, moderate, and severe, each corresponding to different combinations of shortage duration, power gap size, and remaining energy storage support time. Based on this severity assessment results, the system can accurately pinpoint the expected start time of the power supply shortage and predict the expected duration of the shortage event. The time point marking relies on high-precision clock synchronization, while the duration prediction may refer to data from similar historical weather patterns or integrate short-term irradiance forecast information.

[0031] The marked power instability interval is not an isolated time label; it needs to be closely linked to the expected changes in the charging load of parking buildings within that interval. The system obtains the scheduled charging tasks, estimated vehicle arrival times, battery status of currently charging vehicles, and random charging demand predicted based on historical data from the charging pile management system to estimate the expected load increment within the instability interval. The final generated power instability interval is a structured data object containing multi-dimensional attributes. Its core attributes include the duration of the output power decline, the amplitude of the power fluctuation, and the expected power shortage. The duration clarifies the start and end times of the power shortage, the fluctuation amplitude characterizes the severity of the power generation instability, and the expected power shortage quantifies the difference between power generation capacity and load demand. This data object is sent to the subsequent load matching module to provide a time base and power gap basis for determining the priority of power supply. The entire construction process emphasizes the real-time nature of the data and the forward-looking nature of the prediction. The construction of the fluctuation sequence adopts a rolling update mechanism, with new power data constantly being added and the oldest data being removed from the window to maintain the timeliness of the sequence. The threshold setting algorithm possesses self-learning capabilities, slowly adjusting threshold levels based on seasonal changes, holiday load patterns, and other factors to better reflect actual operating conditions. The severity assessment model periodically recalibrates the boundary conditions for level classification using historical operating data, reducing false positives and false negatives. The generation frequency of unstable power supply intervals is consistent with the update cycle of the fluctuation sequence, ensuring the system can respond promptly to rapid changes in power generation conditions. All intermediate calculation results and final generated interval data are timestamped and stored in the database for post-analysis and model optimization. Interval information is published internally via a message bus; subsequent modules subscribing to the message can immediately trigger corresponding processing flows. This loosely coupled design ensures efficient collaboration among system components, jointly addressing the inherent intermittency and volatility challenges of photovoltaic power generation.

[0032] Example 3: See Figure 3The generation of the priority power supply sequence table begins with acquiring the timing information corresponding to the unstable power supply intervals marked by the previous module. This timing information precisely defines the start time, expected duration, and end time of potential power supply shortages, providing a clear time frame for subsequent load matching. Upon receiving this timing frame, the system immediately and synchronously collects real-time operating data from each charging pile within the parking building. This data is acquired through smart meters deployed on each charging pile or monitoring units directly integrated into the charging controller. Specifically, it includes the current load power output of the charging pile and the real-time charging status data of the connected electric vehicle batteries. Battery charging status is typically expressed as a percentage, representing the ratio of current charge to full capacity. Based on this real-time data, the system calculates two key parameters for each active charging task: remaining charging energy and time sensitivity. Remaining charging energy refers to the total electrical energy required to charge the vehicle battery from its current charge to the user's preset target charge or full charge. Its calculation depends on factors such as the battery's current charging status, total battery capacity, and charging efficiency. Time sensitivity is a quantitative indicator that reflects the urgency of the charging task. It is closely related to the user's set expected vehicle departure time. The closer the departure time is, the higher the time sensitivity of the task is usually.

[0033] After calculating the basic parameters of each charging task, the system needs to further incorporate richer contextual information to evaluate its scheduling priority. This information includes the vehicle's scheduled departure time obtained from the parking reservation system and the battery safety charging threshold obtained from the vehicle's battery management system or standard charging protocol. The scheduled departure time provides a baseline for charging completion, while the battery safety charging threshold specifies the maximum allowable charging current or power to avoid overcharging or overheating risks. Assessing the urgency and interruptibility of a charging task is a multi-factor decision-making process. The urgency level is mainly determined by time sensitivity and the remaining energy required for charging. A task with an imminent departure time and a large remaining energy requirement usually has a high urgency level. Interruptibility assesses whether temporarily interrupting or reducing the power supply of the charging task would have an unacceptable impact on battery health, charging completion time, or user agreements. For example, some batteries in the early stages of constant current charging may tolerate larger power fluctuations, while batteries nearing full charge may be more sensitive to power adjustments.

[0034] Based on a comprehensive assessment of the urgency and interruptibility of each charging task, the system assigns them power supply priority numbers and power allocation weights. The power supply priority number is an integer sequence; the smaller the number, the higher the priority, ensuring priority power supply during power shortages. The power allocation weight is a coefficient between 0 and 1, determining how available photovoltaic power resources are specifically allocated to each charging load within that priority level. Assuming the system uses a comprehensive scoring function to aid decision-making, this function can be expressed as:

[0035] in: Representing the The priority score of each charging task is determined by its priority score. A higher score usually means that the priority number assigned to it is smaller (i.e., higher priority). It is a normalized quantification of the urgency of the task; the higher the value, the greater the urgency. This represents the interruptibility assessment value of the task. The closer the value is to 1, the easier it is to be interrupted or have its power reduced. This is the remaining energy required for charging this task. It is the maximum amount of energy remaining to be charged across all active tasks, used for... Normalize. These are the weighting coefficients, which determine the proportion of urgency, interruptibility, and remaining energy in the final score. These coefficients can be configured and adjusted according to the parking lot's management strategy. For example, during peak hours, more emphasis may be placed on urgency, while at night, more emphasis may be placed on overall charging efficiency.

[0036] The final generated priority power supply list is a structured data list, where each record corresponds to an active charging task. Each record contains a unique identifier for the task, its assigned power supply priority number, the calculated power allocation weight coefficient, and the relevant load urgency indicator. Additionally, the priority list may include energy storage compensation activation conditions. These conditions define the circumstances under which the building's energy storage system should be activated to compensate for power supply when photovoltaic power cannot meet the load demand of a certain priority queue. For example, when the total power demand of high-priority loads exceeds a certain proportion of the real-time photovoltaic power generation, energy storage discharge is triggered. The algorithm for generating the priority list needs to have real-time response capabilities to handle dynamic events such as the start of new charging tasks, the completion of existing tasks, or users modifying their departure times. The entire evaluation and sorting process is re-executed in each scheduling cycle to ensure that the power supply strategy can adapt to real-time changes in charging demand within the parking lot. After the priority list is generated, it will be encapsulated into a standard message format and sent to the subsequent power dispatch module through an internal communication interface, serving as the direct basis for its dynamic allocation of photovoltaic power and control of the energy storage system. All raw data, intermediate parameters, and final sorting results used in the calculation process will be recorded for subsequent system performance analysis and optimization iteration of scheduling strategies.

[0037] Example 4: The generation process of photovoltaic power dispatch instructions relies heavily on the load classification information provided by the priority power supply sequence table. At the beginning of a specific dispatch cycle, the system obtains a newly updated priority power supply sequence table from the data bus. This table clearly lists the priority number, power allocation weight, and other relevant attributes of all currently active charging tasks. Assume that at this moment, the output power of the real-time photovoltaic power generation system drops to 35 kW due to cloud cover, while the total load demand of all charging piles in the parking building is calculated to be 50 kW, resulting in a 15 kW power gap that needs to be managed. The core dispatch algorithm first sorts all loads according to the power supply priority number in the sequence table, prioritizing the allocation of limited photovoltaic power resources to the highest priority load group. For example, the load group with priority number 1 will receive power first, until the demand of all loads in that group is met or the photovoltaic power generation capacity is exhausted.

[0038] After power allocation to the highest priority loads is completed, if there is still surplus photovoltaic power, it is recursively allocated to the next lower priority load group. For multiple loads within the same priority group, the system finely allocates the remaining available power according to the power allocation weighting coefficients set in the priority table. The higher the weighting coefficient, the larger the proportion of power allocated to that load. Refer to Table 1, which illustrates the power allocation calculation based on a hypothetical priority power supply table under the current 35 kW photovoltaic power generation capacity. Table 1: Power allocation based on priority power supply order during a certain scheduling cycle

[0039] The system first ensured that the highest priority (number 1) EV-07 charging pile received its full required 10 kW of power. Afterward, the remaining photovoltaic power was 25 kW (35kW - 10kW). Next, the system processed the load group with priority number 2, which included two charging piles, EV-12 and EV-03, with a total requested power of 25 kW, exactly equal to the remaining available power. Based on their power allocation weights (0.6 for EV-12 and 0.4 for EV-03), EV-12 was allocated the following power: However, its requested power was exactly 15 kilowatts, so it was allocated the full amount; the power allocated to EV-03 was... The allocation was initially fully satisfied. However, this allocation result was calculated under the specific scenario of reduced photovoltaic power due to cloud cover. In reality, after allocating EV-07, the remaining power was only 25 kW, while the total requested power for priority group 2 was 25 kW. According to the weighted allocation, EV-12 should have received 15 kW and EV-03 should have received 10 kW, which seems sufficient. However, this example is intended to illustrate the allocation logic. In actual scheduling, photovoltaic power fluctuates in real time, and the allocation algorithm needs continuous adjustment. At the same time, the system must coordinate the management of building energy storage systems to compensate for insufficient photovoltaic power generation or smooth out fluctuations. The scheduling algorithm calculates the power value and duration that the energy storage system needs to release in the current cycle based on the predicted power gap size and duration during the unstable power supply period, as well as the current state of charge of the energy storage system, and adjusts the charging and discharging thresholds of the energy storage converter accordingly. For example, if the prediction shows that the power gap will last for 20 minutes and the energy storage state of charge is sufficient, it may be set to discharge 10 kW of power in the next 20 minutes. The entire power distribution and energy storage dispatch process is under dynamic monitoring. The system tracks the actual fluctuations in photovoltaic power generation, the actual discharge power of the energy storage system, and the actual power consumption of each charging pile in real time. This real-time data is compared with the expected distribution plan. Once a deviation is detected, such as photovoltaic power generation recovering faster than expected, or a high-priority charging task unexpectedly ending, the system will immediately trigger a dynamic correction of the distribution plan, recalculate and issue new power distribution instructions to the relevant charging pile controllers and energy storage converters.

[0040] The final generated photovoltaic power dispatch instruction is a structured set of digital commands. It contains specific control parameters for each controllable object, clearly defining the release time window of the energy storage system—the precise times for starting and stopping discharge—as well as the set discharge power value. For charging loads, the instruction includes a set upper limit or target power value. For some low-priority loads, a load power interruption protection flag may be set, indicating whether a temporary interruption of power supply to that load is permissible in extreme power shortages to protect critical loads. The instruction also defines the range of dynamic power balance, a tolerance band that allows power fluctuations within a certain range, helping to avoid frequent instruction switching due to minor fluctuations. This dispatch instruction is issued to the field actuators via industrial Ethernet or wireless communication networks. The entire generation cycle is very short, typically completed within seconds, ensuring the system can quickly respond to changes in sunlight conditions and maintain a dynamic balance between photovoltaic power supply and demand within the building. All issued instructions, the version of the order table they follow, and real-time system status data are recorded in detail, forming a dispatch log for subsequent analysis, auditing, and algorithm optimization.

[0041] Example 5: The execution of photovoltaic power dispatch instructions triggers a continuous monitoring process for the system's operating status. This process focuses on the actual power supply quality and energy replenishment process of high-priority charging loads. A power quality analyzer deployed on the output side of the charging pile records the voltage and current waveforms transmitted to the vehicle's battery using high-frequency sampling. Power supply stability is quantified by calculating parameters such as effective values ​​and harmonic distortion rates. Simultaneously, the system obtains charging status data from the battery management system via the vehicle communication interface, including charging current received, battery temperature rise rate, and the amount of charge increase per unit time. This data is processed to evaluate the actual battery charging efficiency, with efficiency calculations excluding the influence of auxiliary loads such as battery thermal management energy consumption. Any power interruption event occurring during monitoring is accurately recorded by the system. An interruption event is defined as an unplanned charging power drop to zero and continuously exceeding a set threshold. The recorded content includes the interruption start time, duration, associated charging pile identifier, and the system power balance status immediately before the interruption. The power deviation value is the difference between the average power actually received by the charging pile within the instruction issuance cycle and the target power set by the dispatch instruction. This difference is obtained by comparing the set value with the actual measured value from the meter.

[0042] In-depth analysis of these recorded power outage events and power deviation values ​​is a crucial step in optimizing system decisions. The analysis begins with a time-based approach, examining whether outage events are concentrated in specific time periods, such as during periods of intense afternoon solar radiation fluctuations or evening load peaks, and analyzing their correlation with the simultaneous photovoltaic power generation drop curve. The distribution pattern of power deviation values ​​is characterized using statistical methods. The system calculates the average deviation and standard deviation under different priority loads and different power allocation weights, thereby identifying which allocation strategies are more prone to generating significant errors in actual execution. For example, it may be found that adjacent priority load groups with overly similar weight allocations are more likely to experience command delays during rapid power changes. The analysis also involves cross-comparison, correlating outage events and power deviations with the action records of the energy storage system and the errors in photovoltaic power prediction during the same period, attempting to identify the root causes of power instability. For instance, an outage might be caused by the energy storage system's converter output power ramp-up rate failing to keep up with the photovoltaic power drop rate when responding to dispatch commands.

[0043] Based on the distribution patterns and correlations derived from the analysis, the system initiates revisions to the prediction model for subsequent power supply instability intervals. These revisions may target specific parameters in the model, such as adjusting the threshold sensitivity used to determine the duration of power decline, or modifying the window size of the smoothing algorithm used to predict power shortages. If analysis indicates that historical predictions consistently underestimate the amplitude of power fluctuations under certain weather patterns (such as fast-moving fractulum clouds), the system may add weights for identifying such weather characteristics to the prediction model. The generation logic for the priority power supply order table is also adjusted accordingly, based on retrospective evaluations of historical scheduling effectiveness. For example, if it is found that in a past period, setting the urgency level solely based on vehicle departure times led to unnecessary delays in charging several vehicles with extremely low battery levels, the system may modify the urgency assessment algorithm, introducing the ratio of the current remaining battery charge to the minimum safe driving charge as a new decision factor. Adjustments to the generation logic may also be reflected in the power allocation weight calculation strategy. If statistical analysis shows that simple fixed weight allocation is ineffective when the load dynamically changes, a dynamic weight calculation function based on remaining charging time and energy demand may be introduced.

[0044] Based on the revised prediction model and generation logic, the system ultimately updates the key control parameters included in the photovoltaic power dispatch instructions. The dynamic power balance range is a crucial update target, defining the tolerance range within which actual power deviates from the set value without triggering instruction recalculation. For example, after analyzing historical deviations, the system might adjust the dynamic power balance range of a certain priority load from ±5% to ±8% to adapt to the slow fluctuations in sunlight under thin cloud cover, reducing unnecessary frequent instruction switching. The update of load protection conditions is more detailed, defining the protective actions taken for loads of different priorities under extreme conditions. Corrections may be based on analysis of the impact of historical interruption events. For instance, for some low-priority loads that have been charged to over 95%, the original strategy might have allowed a complete power interruption, but the updated protection conditions might reduce their power to the minimum required for maintenance rather than completely cutting off power, to prevent battery charge from dropping. The entire implementation process of Example 5 constitutes a complete closed-loop optimization system. The execution effect of each scheduling command is transformed into empirical data, which is used to refine each link of the next round of decision-making, so that the adaptability and robustness of the entire power supply management system are continuously improved during continuous operation. All correction operations follow the version management principle. The old and new models and logical versions will coexist for a period of time, and the decision to switch over completely is made by comparing the actual effects, ensuring the smoothness of system optimization. Monitoring data, analysis results, correction parameters, and updated command versions are all recorded in detail, forming a complete audit trail chain.

[0045] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for managing the power supply of photovoltaic parking buildings, characterized in that, Includes the following steps: The real-time irradiance data and module temperature data of the photovoltaic module array on the roof of the parking building are determined. The actual light-receiving area is calculated based on the shadow distribution of the module surface. The photovoltaic output power correction parameters are generated by combining the historical attenuation coefficient. Based on the photovoltaic output power correction parameters, a photovoltaic power generation capacity fluctuation sequence per unit time is constructed, and periods of decreased power generation capacity and power output gaps are identified and marked as unstable power supply intervals. Based on the time-series distribution corresponding to the unstable power supply interval, the real-time load demand data of the charging piles in the parking building is matched, the power difference between the load demand and the power generation capacity is calculated, and a priority power supply order table is generated. According to the sorting results in the priority power supply order table, photovoltaic power is dynamically allocated to charging loads of different priorities, the charging and discharging thresholds of the building energy storage system are adjusted, and photovoltaic power dispatch instructions are generated.

2. The photovoltaic parking building power supply management method according to claim 1, characterized in that, The photovoltaic output power correction parameters include the effective light-receiving ratio of the module, the temperature loss coefficient, and the shading impact factor. The power supply instability range includes the duration of output power decline, the power fluctuation amplitude, and the expected power shortage. The priority power supply order table includes the load urgency level indicator, the power allocation weight coefficient, and the energy storage compensation activation conditions. The photovoltaic power dispatch instructions include the energy storage release time window, the load power supply interruption protection flag, and the power dynamic balance range.

3. The photovoltaic parking building power supply management method according to claim 1, characterized in that, The specific steps for determining the photovoltaic output power correction parameters are as follows: The irradiance and backsheet temperature of the photovoltaic array on the roof of the parking building were collected at multiple sampling times. The shadow distribution pattern on the surface of the array was scanned, and the area ratio and duration of the shadowed area were calculated. Based on the area ratio and duration of the shaded region, combined with the nominal power and temperature characteristic curves of the component, the actual output power loss ratio is calculated, and an irradiation impact feature set is generated. Based on the power loss ratio in the irradiation impact feature set, the historical operating attenuation coefficient of the module and the current environmental cleanliness index are introduced to comprehensively calculate the real-time output capacity correction value of the photovoltaic module and generate photovoltaic output power correction parameters.

4. The photovoltaic parking building power supply management method according to claim 3, characterized in that, The specific steps for constructing the power supply instability range are as follows: The real-time output capability correction value in the photovoltaic output power correction parameters is called to continuously record the fluctuation of photovoltaic power generation over multiple cycles, and to identify the duration and magnitude of the power generation that is below the preset threshold. The severity of the photovoltaic power shortage is determined by statistically analyzing the duration of power generation below a preset threshold and the magnitude of power decline, combined with the current state of charge and dispatchable capacity of the building's energy storage system. Based on the severity of the photovoltaic power supply shortage, the starting point and expected duration of the power supply shortage are marked, and the expected load increase of the charging piles during this period are associated to generate a power supply instability interval.

5. The photovoltaic parking building power supply management method according to claim 4, characterized in that, The specific steps for generating the priority power supply order table are as follows: Obtain the timing information corresponding to the unstable power supply interval, synchronously collect the real-time load power and vehicle battery charging status data of each charging pile in the parking building, and calculate the remaining charging energy and time sensitivity of each charging pile. Based on the remaining charging energy required for each charging station and time sensitivity, combined with the vehicle's scheduled departure time and the battery's safe charging threshold, the urgency and interruptibility of the charging task are assessed. Based on the urgency and interruptibility of the charging task, a power supply priority number and power allocation weight are assigned to each charging load to generate a priority power supply order table.

6. The photovoltaic parking building power supply management method according to claim 5, characterized in that, The specific steps for generating the photovoltaic power dispatch command are as follows: Based on the power supply priority number and power allocation weight in the priority power supply order table, the allocation ratio of photovoltaic power to high-priority charging loads is determined, and the discharge power and duration of the building energy storage system are adjusted at the same time. Monitor the discharge power and duration of the building energy storage system, and dynamically adjust the power supply parameters of different priority loads based on real-time photovoltaic power generation fluctuations. Based on the dynamically adjusted power supply parameters of loads with different priorities, a photovoltaic power dispatch instruction containing power allocation instructions and energy storage control strategies is generated.

7. The photovoltaic parking building power supply management method according to claim 6, characterized in that, The method further includes the following steps: After executing the photovoltaic power dispatch command, continuously monitor the actual power supply stability and battery charging efficiency of high-priority charging loads, and record power interruption events and power deviation values; Analyze the distribution patterns of the power outage events and power deviation values, and revise the prediction model for subsequent unstable power supply intervals and the generation logic of the priority power supply order table. Based on the revised prediction model and generation logic, the power dynamic balance range and load protection conditions in the photovoltaic power dispatch instructions are updated.

8. A photovoltaic parking building power supply management device, used to implement the photovoltaic parking building power supply management method as described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to acquire real-time irradiance data and module temperature data of the photovoltaic module array on the roof of the parking building; The shadow distribution calculation module is used to calculate the actual light-receiving area based on the shadow distribution state of the component surface; The power correction parameter generation module is used to generate photovoltaic output power correction parameters by combining historical attenuation coefficients. A power generation capacity fluctuation sequence construction module is used to construct a photovoltaic power generation capacity fluctuation sequence per unit time based on the photovoltaic output power correction parameters. The power supply instability interval identification module is used to identify periods of reduced power generation capacity and power output gaps and mark them as power supply instability intervals; The load demand matching module is used to match the real-time load demand data of charging piles in the parking building according to the power supply instability range. The power difference calculation module is used to calculate the power difference between load demand and power generation capacity; A priority power supply order table generation module is used to generate a priority power supply order table based on the power difference; The power dispatch module is used to dynamically allocate photovoltaic power to charging loads of different priorities according to the priority power supply order table, adjust the charging and discharging thresholds of the building energy storage system, and generate photovoltaic power dispatch instructions.

9. A photovoltaic parking building power supply management system, 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 photovoltaic parking building power supply management method according to any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the photovoltaic parking building power supply management method according to any one of claims 1 to 7.

Images