A multi-period charging mode scheduling control method for a light storage charging system

Abstract

The application provides a multi-period charging mode scheduling control method for a light storage and charging system, comprising: quantifying the heat dissipation difference characteristics between modules according to the installation positions of the energy storage modules in the cabinet and the cooling air duct structure of the parallel architecture; determining the temperature positions of the module substrates that need to be monitored according to the heat dissipation difference characteristics, collecting the substrate temperature, calculating the change rate of the module substrate temperature with time, and determining the cooling speed of each module; comparing the estimated substrate temperature of each module with the upper limit of the safe working temperature, analyzing the allowed temperature rise space of each module before entering the next working period, and obtaining the temperature rise margin of each module; distributing the expected total current of the photovoltaic output climbing stage to each module according to the temperature rise margin of each module, and identifying the current imbalance risk by judging whether the distributed current of each module matches the temperature rise margin.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for scheduling and controlling multi-period charging modes in a photovoltaic energy storage and charging system. Background Technology

[0002] The scheduling and control of multi-stage charging modes in a photovoltaic-storage-charging system is a core element in ensuring efficient energy utilization and safe equipment operation. In a multi-module parallel operation architecture, modules at the edge of the rack have shorter heat dissipation paths and lower thermal resistance, resulting in faster natural cooling; modules deeper within the rack experience obstructed heat dissipation and higher thermal resistance, leading to slower natural cooling. During the interval between two charging stages, each module cools down naturally at its own rate, and this differentiation in cooling rates widens the temperature difference between modules. The longer the interval, the more residual heat is retained by the slower-cooling modules, and the greater the temperature difference between them and the faster-cooling modules. When the system enters the power ramp-up phase of the next charging stage, this accumulated temperature unevenness during the interval directly translates into an imbalance in current distribution. Since parallel modules share the same bus voltage, modules with higher temperatures have a smaller temperature margin before reaching the thermal derating threshold, and their internal resistance decreases as temperature increases, making them more likely to receive a larger current under the same bus voltage. The combination of these two factors causes these modules to reach their temperature limit first during rapid power ramp-up, thus triggering thermal derating protection prematurely. The derated modules subsequently reduce their own output power share, limiting the total output power of the entire system and preventing them from completing high-power charging tasks according to the scheduling plan. For example, in a photovoltaic-storage-charging integrated power station with multiple parallel energy storage converter modules, module A is installed near the ventilation opening on the front side of the cabinet, while module B is located deep in the rear of the cabinet. After the first round of photovoltaic charging in the morning, the system enters a two-hour interval. During this period, module A's substrate temperature drops from 75°C to 45°C due to its short heat dissipation path and low air resistance; while module B's temperature drops only from 75°C to 62°C due to obstructed airflow and high thermal resistance. When the second round of charging demand arrives in the afternoon, and the system needs to ramp up its output power from zero to the rated value within 30 seconds, all modules start up under the same bus voltage. Since module B currently has a higher temperature, its power device on-resistance is lower, naturally receiving a larger current; at the same time, it only has a 23°C margin from the 85°C thermal derating threshold, far lower than module A's 40°C margin. As a result, when module B's power reached 70%, its temperature rapidly rose to 86°C, triggering a protective derating and forcibly limiting its output. Although the other modules still had some capacity, the overall system output was constrained to 70% of its rated power because the overall control strategy limited the weakest module. This forced an extension of the originally planned 15-minute charging task, impacting the efficiency of the charging station's scheduling and the ability to serve users. Summary of the Invention

[0003] This invention provides a method for scheduling and controlling multi-time-period charging modes in a photovoltaic-storage-charging system, mainly comprising: Based on the installation location and cooling airflow structure of each energy storage module in the rack within the photovoltaic energy storage and charging system, the heat dissipation differences between modules are quantified. Based on these differences, the temperature locations of each module's substrate that need to be monitored are determined, substrate temperatures are collected, and the rate of change of each module's substrate temperature over time is calculated to determine the cooling rate of each module. The interval between adjacent charging periods given by the scheduling plan is received, and the estimated temperature of each module's substrate at the end of the interval is inferred by combining the cooling rate of each module with the current substrate temperature. The estimated substrate temperature of each module is compared with the upper limit of the safe operating temperature to analyze the allowable temperature rise space of each module before entering the next operating period, thus obtaining... Temperature rise margin for each module; allocate the expected total current during the photovoltaic output ramp-up phase according to the temperature rise margin of each module, determine whether the allocated current of each module matches its temperature rise margin, and identify the risk of current imbalance; based on the heat dissipation thermal resistance of the module with current imbalance risk and its allocated current, estimate the temperature rise rate of each module with current imbalance risk when bearing the allocated current, and determine the local overheating probability of each module with current imbalance risk by the degree to which the temperature rise rate exceeds the safe operating temperature rise limit; compare the local overheating probability with a preset probability threshold, obtain the remaining available output margin of the modules whose local overheating probability is lower than the preset probability threshold, and redistribute the output power according to the remaining available output margin.

[0004] Furthermore, based on the installation position and cooling airflow structure of each energy storage module in the rack within the photovoltaic energy storage and charging system, the heat dissipation difference characteristics between each module are quantified. This includes: obtaining the installation position coordinates of each module in the rack, dividing the rack into rows and columns, and recording the distance from each module to the nearest air vent and the upstream and downstream positions of the airflow to form a spatial position file; obtaining the initial thermal resistance value from the substrate of each module to the environment based on the spatial position file, superimposing the thermal coupling resistance value estimated by the heat source density and spacing of adjacent modules to obtain the equivalent heat dissipation thermal resistance and forming a thermal resistance distribution table; calculating the arithmetic mean of the thermal resistance of each module in the thermal resistance distribution table, using the difference between the thermal resistance of each module and the average value as a heat dissipation deviation index, and combining the airflow path to identify fast-cooling modules and slow-cooling modules, and incorporating them into the heat dissipation difference characteristic file.

[0005] Furthermore, the step of determining the temperature location to be monitored for each module substrate based on the heat dissipation difference characteristics, collecting substrate temperature, calculating the rate of change of substrate temperature over time for each module, and determining the cooling rate of each module includes: determining measurement points by placing points at the projection position directly above the power device of each module and at the edge position of the heat dissipation substrate according to the fast and slow cooling module labels in the heat dissipation difference characteristic file; attaching thermistor elements at the measurement points to collect substrate temperature at a preset sampling rate to obtain a base temperature sequence, and taking the absolute value of the ratio of the temperature difference between adjacent time windows to the time window duration as the cooling rate of each module after sliding time window mean filtering, and archiving it according to module number.

[0006] Furthermore, the method of receiving the interval duration of adjacent charging periods given by the scheduling plan, and inferring the estimated temperature of each module's substrate at the end of the interval duration by combining the cooling rate of each module with the current substrate temperature, includes: dividing the interval duration into a time step sequence according to a preset step size, collecting the ambient temperature inside the cabinet as a reference value for the final cooling temperature; establishing a time-step temperature decay recursive relationship based on the cooling rate of each module and the difference between the current substrate temperature and the reference value, accumulating the temperature drop along the time step sequence to obtain the temperature drop trajectory, and archiving the end temperature value corresponding to the end of the interval to form an estimated temperature file.

[0007] Furthermore, comparing the estimated substrate temperature of each module with the upper limit of the safe operating temperature, and analyzing the allowable temperature rise space of each module before entering the next working period to obtain the temperature rise margin of each module, includes: obtaining the estimated temperature of the substrate of each module from the estimated temperature file; adjusting the initial upper limit of the safe operating temperature calibrated at the factory of the module according to the heat dissipation deviation index to obtain the upper limit value; taking the difference between the estimated temperature and the upper limit value as the temperature rise space of each module; adding a lenient label to the module whose temperature rise space is higher than the preset warning threshold, and adding an urgent label to the module whose temperature rise space is lower than the preset warning threshold, and importing them into the temperature rise margin file.

[0008] Furthermore, the process of allocating the expected total current for the photovoltaic output ramp-up phase according to the temperature rise margin of each module, determining whether the allocated current for each module matches its temperature rise margin, and identifying current imbalance risks includes: using the proportion of each module's temperature rise space in the total as the allocation weight, proportionally dividing the expected total current for the ramp-up phase issued by the scheduling plan according to the allocation weight to obtain the allocated current for each module; using the square of the allocated current and the product of the module's equivalent heat dissipation thermal resistance as the temperature rise estimate, subtracting the corresponding temperature rise space from the temperature rise estimate to obtain the matching difference; attaching an imbalance flag to modules with a matching difference higher than a preset imbalance threshold and marking them as having current imbalance risk, and attaching a normal flag to modules with a matching difference lower than the imbalance threshold, and merging the module number and the corresponding flag into the imbalance risk file.

[0009] Furthermore, the step of estimating the temperature rise rate of each module at risk of current imbalance when bearing the distributed current, based on the heat dissipation thermal resistance of the module with current imbalance risk and its distributed current, includes: obtaining the module number with imbalance identifier and the corresponding distributed current; retrieving the equivalent heat dissipation thermal resistance from the thermal resistance distribution table for each imbalanced module and reading the on-resistance and thermal capacity time constant of the power device; using the product of the square of the distributed current and the on-resistance as the Joule power; and dividing the product of the Joule power and the equivalent heat dissipation thermal resistance by the thermal capacity time constant to obtain the temperature rise rate of each imbalanced module.

[0010] Furthermore, determining the probability of local overheating of each module with current imbalance risk based on the degree to which the temperature rise rate exceeds the safe operating temperature rise limit includes: retrieving the safe operating temperature rise limit for each imbalance module, adjusting the safe operating temperature rise limit downward based on the heat dissipation deviation index to obtain a safe temperature rise benchmark; subtracting the safe temperature rise benchmark from the temperature rise rate to obtain the temperature rise exceeding the limit, dividing the temperature rise exceeding the limit by the safe temperature rise benchmark to obtain an exceeding degree coefficient; and performing a nonlinear transformation on the exceeding degree coefficient using a probability mapping based on the logistic function to obtain a probability value mapped to between 0 and 1, which serves as the probability of local overheating for each module with current imbalance risk.

[0011] Furthermore, the step of comparing the local overheating probability with a preset probability threshold to obtain the remaining usable output margin of modules whose local overheating probability is lower than the preset probability threshold includes: comparing the local overheating probability of each module with the preset probability threshold, marking modules with a probability higher than the preset probability threshold as restricted modules, and classifying modules with a probability lower than or equal to the preset probability threshold into a safe module set; for each module in the safe module set, multiplying the temperature rise space by the heat capacity time constant and then dividing by the product of the equivalent heat dissipation thermal resistance and the climb-up stage duration to obtain the remaining usable output margin.

[0012] Furthermore, after obtaining the remaining available output margin, the process includes: summing the remaining available output margins of each module in the set of safety modules to obtain the total power that the set can carry; receiving the total power command during the ramp-up phase, wherein if the total power command exceeds the total power that the set can carry, the power is truncated to the total power that the set can carry; proportionally dividing the obtained total power according to the proportion of the remaining available output margin of each module to obtain the redistributed output power; dividing the redistributed output power by the bus voltage sampling value of each safety module during the power ramp-up phase to obtain the target distribution current of the module; and setting the target distribution current to zero for the restricted modules.

[0013] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a multi-period charging mode scheduling and control method for a photovoltaic-storage-charging system. Addressing the problem of uneven power distribution and localized overheating risks caused by varying heat dissipation conditions due to different installation positions of energy storage modules within the cabinet, this invention proposes a dynamic current allocation strategy based on differences in thermal characteristics. By quantifying the differences in thermal resistance of each module, monitoring the substrate temperature change rate in real time to calculate the cooling rate, and combining the charging period intervals to infer the estimated temperature of each module at the start of the next working period, this invention analyzes the allowable temperature rise space of each module to obtain a temperature rise margin. During the photovoltaic output ramp-up phase, this invention allocates current based on the temperature rise margin and identifies imbalance risks. For modules with risks, the invention estimates their temperature rise rate and quantifies the probability of localized overheating. By comparing probability thresholds, the power allocation scheme is dynamically adjusted to transfer the load to modules with sufficient temperature rise margin, ultimately determining the target allocation current for each module. This method effectively avoids thermal imbalance problems caused by differences in heat dissipation conditions, improving system safety and overall output capacity. Attached Figure Description

[0014] Figure 1 This is a flowchart of a multi-period charging mode scheduling and control method for a photovoltaic energy storage and charging system according to the present invention.

[0015] Figure 2 This is a schematic diagram of a multi-period charging mode scheduling and control method for a photovoltaic energy storage and charging system according to the present invention.

[0016] Figure 3 This is another schematic diagram of a multi-period charging mode scheduling and control method for a photovoltaic energy storage and charging system according to the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] like Figures 1-3 This embodiment of a multi-period charging mode scheduling and control method for a photovoltaic energy storage and charging system may specifically include: S101. Based on the installation position of each energy storage module in the cabinet and the cooling air duct structure of its parallel architecture in the photovoltaic energy storage and charging system, quantify the heat dissipation differences between the modules.

[0019] For parallel energy storage modules within a photovoltaic-energy storage-charging system cabinet, the coordinates of each module's installation location within the cabinet are obtained. The cabinet is then partitioned according to row and column levels, resulting in a partition number for each module. Based on the arrangement of the cooling fan inlets and outlets within the cabinet, airflow paths are plotted, recording the distance from each module to the nearest air outlet and its relative upstream or downstream position in the airflow. The module numbers, partition numbers, air outlet distances, and airflow path attributes are merged to form a spatial location file. Based on this spatial location file, the initial thermal resistance between the substrate and the environment is obtained for each module using steady-state heating measurements or manufacturer calibration data. Then, based on the heat source density and spacing of adjacent modules, the thermal coupling resistance between modules is estimated. For modules located downstream of the airflow or deep within the cabinet, the thermal coupling resistance is superimposed on the initial thermal resistance to obtain the equivalent heat dissipation thermal resistance. The equivalent heat dissipation thermal resistance of each module is archived according to its module number, forming a thermal resistance distribution table. Based on the thermal resistance distribution table, the arithmetic mean of the equivalent heat dissipation thermal resistance of each module is calculated. The difference between the thermal resistance of each module and the average value is used to obtain the heat dissipation deviation index. Combining the airflow path and the distance to the air outlet, the deviation index is classified and labeled as upstream, midstream, and downstream, identifying fast cooling modules with low thermal resistance and slow cooling modules with high thermal resistance. The deviation index, partition number, and thermal resistance value of the fast cooling module and the slow cooling module are entered into the heat dissipation difference feature file to obtain the heat dissipation difference characteristics between each module.

[0020] In the actual deployment of integrated photovoltaic, energy storage and charging power stations, parallel energy storage modules are usually installed in the same cabinet in multiple rows and multiple layers. Air inlets are arranged on the front and top of the cabinet, and air outlets are arranged on the rear or bottom.

[0021] In one possible implementation, the cabinet is calibrated in three dimensions. The origin is the lower left corner of the front panel, and coordinate axes are established along the horizontal, vertical, and depth directions. The coordinate values ​​of the geometric center of each module are recorded. Alternatively, based on the arrangement of the air inlets and outlets of the cooling fans within the cabinet, the airflow path can be plotted using smoke tracing or computational fluid dynamics simulation. Specifically, tracer smoke is released at the air inlet or a three-dimensional flow field model is established. The trajectory of the smoke flowing across the surface of each module is observed or calculated, marking the main airflow channel and branch paths. The straight-line distance from the center point of each module to the nearest air outlet edge is recorded as the air outlet distance D. Modules with D less than 150 mm are marked as upstream modules, those with D between 150 and 300 mm are marked as midstream modules, and those with D greater than 300 mm are marked as downstream modules.

[0022] Specifically, the spatial location archive is established by the following method: numbering the cabinets by column, by layer, and by depth, and attaching a partition number consisting of column number, layer number, and depth number to each module.

[0023] For example, modules near the air inlet on the front are assigned to the shallow zone, modules in the middle of the rack to the middle zone, and modules near the rear baffle to the deep zone. The orientation of the heat sink substrate, the spacing between adjacent modules, and the projected positions of the power devices in adjacent modules above and below are also recorded. The airflow path is determined by the arrangement of fans and the baffle structure inside the rack.

[0024] In one embodiment, the mainstream direction of the cooling airflow is obtained by using smoke tracing or computational fluid dynamics simulation, the upstream, midstream or downstream position of each module in the airflow is marked, and the straight-line distance from the power device of each module to the nearest air inlet is measured.

[0025] Preferably, for areas with cable obstructions or filter blockages, local blockage markers are recorded. When obtaining the initial thermal resistance value, under stable ambient temperature conditions in the cabinet, a constant heating power P is applied to the power device of a single module under test. After the temperature reaches a steady state, the substrate temperature Tb is measured using a thermistor embedded in the substrate, and simultaneously, the ambient temperature Ta at the same height within the cabinet is measured. The initial thermal resistance value of the module is obtained according to R0 = (Tb - Ta) / P, where Tb is in Kelvin, Ta is in Kelvin, P is in watts, and R0 is in Kelvin per watt. This process is repeated for each module to obtain a sequence of initial thermal resistance values ​​corresponding to each module number. The thermal coupling resistance reflects the equivalent thermal resistance increase caused by mutual heating between adjacent modules.

[0026] In one embodiment, based on the heat source density of the power devices in adjacent modules (i.e., the heat generation power per unit heat dissipation area) and the geometric distance between the two modules, an additional thermal coupling resistance is applied to modules located in areas of high heat source density or downstream of the airflow, and this resistance is added to the initial thermal resistance value. Further, the initial thermal resistance value is added to the thermal coupling resistance value to obtain the equivalent heat dissipation thermal resistance, denoted as Req. The equivalent heat dissipation thermal resistances of all modules are entered into a thermal resistance distribution table in order of module number and zone number. Each row of the thermal resistance distribution table corresponds to one module, and the column fields include module number, zone number, air outlet distance, and Req value. After the thermal resistance distribution table is completed, the arithmetic mean Rm of the Req values ​​of all modules is calculated. The difference between Req and Rm for each module is taken as the heat dissipation deviation index.

[0027] For example, if the Req of a shallow partitioning module is less than Rm, the difference is negative, and it is identified as a fast cooling module; if the Req of a deep partitioning module is greater than Rm, the difference is positive, and it is identified as a slow cooling module.

[0028] Understandably, the larger the absolute value of the difference, the more the module's heat dissipation performance deviates from the overall average level of the server rack.

[0029] Preferably, the deviation indicators are categorized into upstream, midstream, and downstream subsets for easy retrieval based on airflow location groups. The final heat dissipation difference feature file includes the following fields: module number, zone number, vent distance, equivalent thermal resistance, heat dissipation deviation index, and fast / slow cooling label. This file provides a module-level heat dissipation baseline for the photovoltaic-storage-charging system during multi-time period switching, intermittent cooling, and power ramp-up phases, avoiding the crude judgments caused by relying on a single temperature threshold.

[0030] S102. Based on the heat dissipation difference characteristics, determine the temperature location that needs to be monitored for each module substrate, collect the substrate temperature, calculate the rate of change of the substrate temperature of each module over time, and determine the cooling rate of each module.

[0031] Starting from the heat dissipation difference characteristic file, and based on the fast and slow cooling module labels and thermal resistance distribution, measurement points are determined on the substrate by placing points at the projection position directly above the power device of each module and at the edge of the heat dissipation substrate. Thermistors are attached to these measurement points, and the substrate temperature is continuously collected at a preset sampling rate to obtain a base temperature sequence aligned with the time stamp. The base temperature sequence is then subjected to mean filtering within a sliding time window to remove instantaneous interference, resulting in a smoothed base temperature curve. The rate of change of the substrate temperature is obtained by the ratio of the temperature difference between adjacent time windows to the time window duration. The absolute value of this rate of change is taken as the cooling rate of each module, and the modules are archived according to their module numbers to determine the cooling rate of each module.

[0032] In one possible implementation, measurement points are arranged for each parallel energy storage module based on the fast and slow cooling module labels and thermal resistance distribution given in the heat dissipation difference feature file.

[0033] Specifically, a main measurement point is selected at the projection position directly above the power device, and an auxiliary measurement point is selected at the edge of the heat dissipation substrate near the cabinet wall. For modules with high thermal resistance and marked as slow cooling, an additional supplementary measurement point is added in addition to the main measurement point, which is close to the output terminal of the power device, to reflect the actual substrate temperature in areas with steep temperature gradients.

[0034] It should be noted that a negative temperature coefficient thermistor is attached to the measuring point, and the lead wire is connected to the acquisition circuit through an electrical isolation plate.

[0035] In one embodiment, the acquisition loop continuously samples the substrate temperature at each measurement point at a sampling rate of 1 Hz, and timestamps each sample value according to a unified time base to obtain a base temperature sequence aligned with the time stamp. Further, the base temperature sequence is subjected to mean filtering within a 30-second sliding time window, with the time window advancing by 5 seconds per step, to obtain a smoothed base temperature curve.

[0036] For example, by taking the difference between the average temperatures in two adjacent time windows and dividing it by the duration between the center moments of the two time windows, we can obtain the rate of change of the substrate temperature of the module in the current time period, denoted as v.

[0037] Preferably, samples in the natural cooling phase during the charging interval are screened, and the absolute value of the rate of change |v| is taken as the cooling rate of the module.

[0038] It is understandable that modules installed near the front ventilation openings of the rack tend to have a larger |v|, while modules located deep in the rear of the rack have a smaller |v|. The resulting cooling rates of each module are archived according to their module numbers, corresponding to the fast and slow cooling labels.

[0039] S103. Receive the interval duration of adjacent charging periods given by the scheduling plan, and infer the estimated temperature of each module's substrate at the end of the given interval duration of adjacent charging periods by combining the cooling rate of each module with the current substrate temperature.

[0040] The interval duration between two adjacent charging periods is obtained from the scheduling order. The interval duration is divided into a time-step sequence according to a preset step size. The current substrate temperature of each module is obtained as the inference starting point, and the ambient temperature at the corresponding location in the cabinet is collected as the reference value for the final cooling temperature. Based on the cooling rate of each module and the difference between the current substrate temperature and the reference final temperature, a time-step temperature decay recursive relationship is established. The temperature drop in each time step is given by the product of the cooling rate and the time step duration. The current temperature difference is updated with the starting temperature of the new time step to obtain the time-step temperature drop along the time-step sequence. The temperature drop is accumulated hourly from the starting point of the time-step sequence, and the temperature drop trajectory of each module substrate is inferred forward along the interval. The end temperature value corresponding to the drop trajectory at the end of the interval is taken and archived according to the module number to form an estimated temperature file, thus obtaining the estimated temperature of each module substrate.

[0041] In the daily operation of an integrated photovoltaic, energy storage, and charging power station, dispatch orders are usually issued by the station's energy management backend, specifying the start and end times of each charging task for the day, as well as the interval between tasks.

[0042] Specifically, the period from the end of the last charging task to the start of the next charging task is called the interval, and its duration is read in minutes.

[0043] In one possible implementation, the duration of the interval is divided into a sequentially arranged time step sequence according to a preset step size.

[0044] For example, if the preset step size is 10 seconds and the interval is 120 minutes, the time step sequence contains a total of 720 time steps, and each time step corresponds to an inference time point.

[0045] Preferably, for scenarios with short intervals, the preset step size can be reduced to 5 seconds to improve the resolution of the trajectory.

[0046] It should be noted that the final temperature reference value represents the temperature that the substrate will approach after the module has cooled naturally for a sufficiently long time.

[0047] Specifically, the ambient temperature of each module in the cabinet is collected, including the incoming air temperature near the air inlet and the steady-state temperature of different zones inside the cabinet. The ambient temperature corresponding to each module is taken as the final temperature reference value of the module according to its zone number.

[0048] Understandably, the final temperature reference value for modules closer to the air inlet is closer to room temperature, while the final temperature reference value for modules deeper in the cabinet is often slightly higher. After determining the time step sequence and the final temperature reference value for each module, a time-step recursive relationship for temperature decay is established.

[0049] In one embodiment, the recursive relationship adopts a first-order discrete form, T(k+1) = T(k) - v·Δt, where T(k) represents the temperature of the module substrate at the starting point of the k-th time step, T(k+1) represents the temperature of the module at the starting point of the (k+1)-th time step, v represents the cooling rate of the module, and Δt represents the preset step size. After each time step calculation is completed, T(k) is replaced by T(k+1) as the current temperature of the next time step, thereby obtaining the time-step temperature drop sequence.

[0050] Preferably, when T(k) has approached the final temperature reference value, the temperature drop of this time step and its subsequent time steps is limited to zero to avoid the temperature from deviating negatively from the reference value.

[0051] Specifically, let the final state temperature reference value be Ttarget, when the following conditions are met... At a temperature of [temperature value] degrees Celsius, it is determined to be close to the final state temperature reference value, where T(k) is the temperature value at the k-th time step, and the vertical line indicates taking the absolute value. At this time, the temperature drop ΔT at the k-th time step and all subsequent time steps is set to 0 degrees Celsius, so that the temperature remains at the current value and does not decrease further, thereby ensuring that the temperature does not deviate negatively from Ttarget, and guaranteeing that the temperature is controlled within ±0.5 degrees Celsius of the reference value. In another embodiment, according to the cooling rate K of each module, the difference between the current substrate temperature T0 and the final state temperature reference value Te is [temperature value]. Establish a time-step temperature decay recursive relationship, wherein the temperature drop in the nth time step is... , where Δt is the time step duration, ∑ΔTᵢ is the cumulative temperature drop in the first n-1 steps, and the current temperature difference is updated with the starting temperature of the new time step to obtain the step-by-step temperature drop along the time step sequence.

[0052] Understandably, the defined cooling rate already reflects the heat dissipation conditions of the module in the rack.

[0053] For example, modules installed near the front ventilation openings cool down faster, resulting in a larger hourly temperature drop; modules located deep in the rear of the cabinet cool down slower, resulting in a smaller hourly temperature drop. Furthermore, the temperature drop is accumulated hourly from the first to the last time step of the time step sequence, forming a temperature drop trajectory for each module substrate within the interval. This trajectory, with the time step number on the x-axis and the calculated substrate temperature on the y-axis, reflects the process of the module substrate gradually approaching the final temperature reference value from its current substrate temperature.

[0054] For example, the end of the interval corresponds to the end of the last time step in the time step sequence. The temperature value corresponding to this end position is read on the descent trajectory, which is the end temperature value of the module at the end of the interval. This reading process is repeated for all parallel modules in sequence, and the estimated temperature files are formed by archiving according to the module number and the cabinet partition to which it belongs.

[0055] Understandably, the estimated temperature profile includes both the lower end temperature values ​​of modules near the front vents and the higher end temperature values ​​of modules deep inside the cabinet, reflecting the actual temperature distribution of multiple modules after cooling and differentiation during the interval period until the start of the next charging task, thus obtaining the estimated temperature of each module's substrate.

[0056] S104. Compare the estimated substrate temperature of each module with the upper limit of the safe operating temperature, analyze the allowable temperature rise space of each module before entering the next working period, and obtain the temperature rise margin of each module.

[0057] The estimated temperature of each module substrate is obtained from the estimated temperature file. The initial upper limit of the safe operating temperature specified in the factory specifications of each module is retrieved. The initial upper limit is adjusted downward according to the heat dissipation deviation index in the heat dissipation difference characteristic file to obtain the upper limit value. The estimated temperature of each module is compared with the upper limit value one by one, and the temperature difference between the two is calculated. The temperature difference is used as the allowable temperature rise space for each module before entering the next working period. A leeway label is attached to the module whose temperature rise space is higher than the preset warning threshold, and an urgent label is attached to the module whose temperature rise space is lower than the warning threshold. The module number, the temperature rise space, and the corresponding label are archived together to form a temperature rise margin file to obtain the temperature rise margin of each module.

[0058] In one possible implementation, the initial upper limit of the safe operating temperature comes from the module's manufacturer's specifications, and is usually reflected as the substrate temperature derating threshold, with a common value in the 85-degree Celsius range.

[0059] It should be noted that modules located deep within the cabinet or with large heat dissipation deviations, which are subjected to high steady-state temperatures and weak airflow over long periods, need to have their actual tolerable temperature adjusted downwards relative to the initial upper limit.

[0060] Specifically, based on the heat dissipation deviation index of the corresponding module in the heat dissipation difference feature file, a downward adjustment amount between 0 and 5 degrees Celsius is linearly mapped according to the magnitude of the absolute value of the deviation index. This amount is then subtracted from the initial upper limit to obtain the upper limit value of the module. Optionally, the downward adjustment amount is calculated according to the following rules: the absolute value of the heat dissipation deviation index of each module is normalized across all modules. When the normalization value is 0, the downward adjustment amount is 0 degrees Celsius; when the normalization value is 1, the downward adjustment amount is 5 degrees Celsius; intermediate values ​​are linearly interpolated proportionally. This downward adjustment amount is then subtracted from the initial upper limit to obtain the upper limit value of the module. The estimated temperature file is generated by the predicted temperature calculation of each module in step S103 and includes the estimated temperature value of each module. Further, the estimated temperature of each module in the estimated temperature file is compared with the upper limit value one by one, according to... The temperature difference is calculated, where Tlim represents the upper limit, Tpred represents the estimated temperature, and ΔT represents the temperature difference between the two.

[0061] It is understood that the temperature difference represents the temperature rise that the module can still tolerate before entering the next charging period.

[0062] For example, modules near the front ventilation openings of the cabinet are expected to have lower temperatures and greater potential for temperature rise; modules deep in the rear of the cabinet are expected to have higher temperatures and less potential for temperature rise.

[0063] Preferably, the preset warning threshold is 15 degrees Celsius. Modules with a temperature rise margin higher than this threshold are labeled with a leeway tag, and modules with a temperature rise margin lower than this threshold are labeled with a urgency tag. The module number, temperature rise margin value, and leeway or urgency tag are written into the temperature rise margin file in rows to obtain the temperature rise margin of each module, which can be called during the power ramp-up phase.

[0064] S105. Allocate the expected total current for the photovoltaic output ramp-up phase to each module according to the temperature rise margin of each module, and identify the risk of current imbalance by judging whether the current allocated to each module matches its temperature rise margin.

[0065] The temperature rise margin of each module is obtained from the temperature rise margin file; the proportion of each module's temperature rise margin in the sum of all module temperature rise margins is used as the allocation weight; the expected total current of the photovoltaic output ramp-up phase issued by the scheduling plan is received, and the expected total current is proportionally divided according to the allocation weight to obtain the allocated current of each module. Based on the allocated current of each module, the product of the square of the current and the equivalent heat dissipation thermal resistance of the module is used as the estimated temperature rise caused by the allocated current; for each module, the difference between the estimated temperature rise and the corresponding temperature rise margin is taken as the matching difference reflecting the degree of deviation of the allocated current from the temperature rise margin; the matching differences are arranged into a matching difference sequence according to the module number. According to the matching difference sequence, modules with matching differences higher than the preset imbalance threshold are marked with an imbalance label, indicating that they have a current imbalance risk; modules with matching differences lower than the imbalance threshold are marked with a normal label; the module number, matching difference, and corresponding label are merged into the imbalance risk file to obtain the current imbalance risk identification result.

[0066] When the photovoltaic-storage-charging system enters the next charging period, the dispatch order synchronously sends a power ramp-up command to each parallel energy storage module, corresponding to an expected total current Itot. The expected total current is derived by the power station energy management backend based on the photovoltaic output curve and load demand, and the unit is usually amperes.

[0067] It should be noted that the expected total current is not evenly distributed among the modules, but rather proportionally divided based on the remaining temperature rise space of each module before it enters the ramp-up phase.

[0068] Specifically, for a cabinet containing n parallel modules, the temperature rise space of each module is denoted as ΔTi (i=1,2,...,n), where ΔTi represents the allowable temperature rise value of the i-th module before entering the next working period.

[0069] In one possible implementation, the proportion of ΔTi in the sum of the temperature rise space of all modules is used as the allocation weight wi of that module, according to... To obtain.

[0070] For example, in a cabinet containing four modules with temperature rise spaces of 40, 35, 25, and 12 degrees Celsius, the allocation weights for each module are approximately 0.345, 0.302, 0.216, and 0.103, respectively. It can be seen that modules with larger temperature rise spaces receive greater weights, while modules with smaller temperature rise spaces receive smaller weights, with the sum of the allocation weights always being 1. Further, the expected total current is proportionally divided according to the allocation weights of each module, resulting in the allocated current for each module: Ii = wi × Itot. Using the previous example of four modules, when the expected total current is 200 amps, the allocated currents for each module are approximately 69.0, 60.4, 43.2, and 20.6 amps, respectively, exhibiting a current gradient consistent with the temperature rise space distribution. After determining the allocated current, a temperature rise prediction is performed for each module. The predicted temperature rise is the amount of temperature increase caused by the allocated current flowing through the module during the ramp-up period.

[0071] In one embodiment, the temperature rise estimate In this formula, ΔTpred,i represents the estimated temperature rise of the i-th module, Ii represents the current value allocated to the i-th module, Req,i represents the equivalent thermal resistance of the i-th module, and Krated represents the normalization coefficient of the square of the current with respect to the temperature rise under rated operating conditions. This coefficient is derived from factory thermal model tests and typically ranges from 0.8 to 1.2. The formula calculates the temperature rise corresponding to power loss by multiplying the square of the current by the equivalent thermal resistance. The Krated coefficient is used to correct for the difference between actual and rated operating conditions.

[0072] It is understandable that modules with smaller temperature rise margins and higher thermal resistance will have more significant temperature rise estimates when handling the same proportion of current.

[0073] It should be noted that the matching difference reflects the relative relationship between the estimated temperature rise caused by the distributed current and the allowable temperature rise range of the module.

[0074] Specifically, for each module, di = ΔTpred,i - ΔTi is taken as the matching difference. If di is negative or close to zero, it indicates that the temperature rise caused by the distributed current is still within the allowable temperature rise range; if di is positive, it indicates that the temperature rise caused by the distributed current exceeds the allowable temperature rise range. The matching differences are arranged sequentially according to the module number to form a matching difference sequence. A preset imbalance threshold dlim is used to distinguish between normal and imbalanced categories.

[0075] For example, dlim = 2 degrees Celsius. Modules with a matching difference higher than the imbalance threshold are marked with an imbalance flag, indicating a risk of current imbalance; modules with a matching difference lower than the imbalance threshold are marked with a normal flag, indicating that their current distribution is within the allowable range. The module number, allocated current, estimated temperature rise, matching difference, and corresponding flag for each module are written into the imbalance risk file row by row to obtain the current imbalance risk identification result.

[0076] Preferably, the imbalance markers in the imbalance risk file are arranged from high to low according to the matching difference, so that the module with the most serious temperature rise can be checked first during subsequent verification.

[0077] It is understandable that the current imbalance risk identification result is directly related to the current sharing relationship of each module in the cabinet after the accumulated temperature unevenness during the interval period and during the power ramp-up phase, providing a preliminary criterion for whether the output power needs to be reallocated in the next period. Another way to identify the current imbalance risk is to allocate the expected total current of the photovoltaic output ramp-up phase to each module according to the temperature rise margin of each module, and to determine whether the allocated current of each module matches its temperature rise margin: obtain the temperature rise space value ΔTi of each module from the temperature rise margin file, and calculate the allocation weight wi of each module according to the formula wi=ΔTi / ΣΔTj; receive the expected total current Itotal of the photovoltaic output ramp-up phase issued by the scheduling plan, and calculate the allocated current of each module according to the formula Ii=wi×Itotal. Based on the allocated current Ii of each module, according to the formula... Calculate the steady-state temperature rise estimate, where Reqi is the equivalent thermal resistance of the module; for each module, subtract the corresponding temperature rise range from the estimated temperature rise estimate. The matching difference is used to reflect the degree to which the distributed current deviates from the temperature rise margin. The matching differences are arranged into a matching difference sequence according to the module number. Based on the matching difference sequence, modules with matching differences higher than a preset imbalance threshold of 5 degrees Celsius are marked with an imbalance flag, indicating that they have a current imbalance risk; modules with matching differences lower than the imbalance threshold are marked with a normal flag; the module number, matching difference, and corresponding flag are entered into the imbalance risk file to obtain the current imbalance risk identification result.

[0078] S106. Based on the thermal resistance of the module with current imbalance risk and its distributed current, estimate the temperature rise rate of each module with current imbalance risk when it bears the distributed current, and determine the local overheating probability of each module with current imbalance risk by the degree to which the temperature rise rate exceeds the safe operating temperature rise limit.

[0079] Obtain the module number with imbalance identifier and corresponding allocated current from the imbalance risk file; for each imbalanced module, retrieve the equivalent heat dissipation thermal resistance from the thermal resistance distribution table, and read the on-resistance and thermal capacity time constant of the power device; use the product of the square of the allocated current and the on-resistance as the Joule power, and then divide the product of the Joule power and the equivalent heat dissipation thermal resistance by the thermal capacity time constant to obtain the temperature rise rate of each imbalanced module. For each imbalanced module, retrieve the safe operating temperature rise limit, and adjust the safe operating temperature rise limit downwards according to the heat dissipation deviation index to obtain the safe temperature rise benchmark; subtract the safe temperature rise benchmark from the temperature rise rate to obtain the temperature rise exceeding the limit; divide the temperature rise exceeding the limit by the safe temperature rise benchmark to obtain the exceedance degree coefficient. Based on the exceedance degree coefficient, use a probability mapping based on the logistic function to perform a nonlinear transformation on the exceedance degree coefficient, mapping the exceedance degree coefficient to a probability value between 0 and 1, to obtain the local overheating probability of each module with current imbalance risk.

[0080] This solution is applied to an integrated photovoltaic-energy storage-charging station system. This system integrates photovoltaic power generation, energy storage devices, and charging facilities to provide charging services for electric vehicles. The system operates in cycles based on working periods, with each working period corresponding to a charging period. At the start of a charging period, the system initiates a power ramp-up phase to increase the output power from standby to the target charging power. When the system enters the next charging period and initiates the power ramp-up, the controller filters all modules with imbalance indicators from the imbalance risk file, obtaining a list of module numbers and corresponding current allocations. The scheduling cycle continues at preset intervals, with a dynamic adjustment to the current allocation during the ramp-up phase within each scheduling cycle. The ramp-up phase lasts for several consecutive scheduling cycles until a steady-state charging power is reached.

[0081] It should be noted that this section only focuses on modules marked as having current imbalance; modules marked as normal are not included in the temperature rise rate calculation in this step.

[0082] Specifically, for each unbalanced module in the list, the equivalent thermal resistance Req of that module is read from the thermal resistance distribution table. This parameter represents the total thermal resistance from the power device node to the ambient temperature, and is the sum of the thermal resistance from the chip inside the module to the substrate, the contact thermal resistance from the substrate to the heat sink, and the convective thermal resistance from the heat sink to the environment, expressed in degrees Celsius per watt. For forced air cooling systems, the front module has a typical thermal resistance of 0.4 to 0.6 due to the low airflow temperature and high flow rate; the middle module has a typical thermal resistance of 0.7 to 0.9 due to airflow preheating and decreased flow rate; and the rear module has a typical thermal resistance of 1.0 to 1.3 due to the cumulative temperature rise from the airflow. The on-resistance Ron of the power device is read from the module's factory parameter table, expressed in milliohms. The thermal capacity time constant τ is read from the same parameter table. This parameter is defined as the time required for the temperature rise to reach 63.2% of the steady-state value under a step heating power, and is physically equal to the product of the module's total heat capacity C and the equivalent thermal resistance Req, expressed in seconds. For a copper-based power module, if the substrate mass is m kg and the specific heat capacity c is 0.385 kJ / kg / degree Celsius, then the heat capacity C equals m multiplied by c multiplied by 1000. Combining this with the aforementioned Req, we can obtain τ. When the substrate mass is 0.2 kg and Req is 0.8 degrees Celsius / watt-hour, τ is approximately 60 seconds. The heat capacity time constant reflects the thermal inertia of the module substrate and the heat dissipation structure; a larger value indicates a slower temperature response under the same heat output. Joule power reflects the instantaneous heat generated when the distributed current flows through the power device.

[0083] In one possible implementation, the Joule power Pj is obtained by multiplying the square of the distribution current by the on-resistance, written as Pj=I²×Ron, where I represents the distribution current of the imbalance module.

[0084] For example, if the distribution current of an imbalance module is 80 amperes and the on-resistance is 5 milliohms, then the Joule power is approximately 32 watts. Furthermore, the temperature rise rate reflects how quickly the substrate temperature rises over time under continuous Joule power input. The temperature rise rate r is given by the following relationship: r = Pj × Req / τ, where Pj represents the aforementioned Joule power, Req represents the aforementioned equivalent thermal resistance, τ represents the aforementioned thermal capacity time constant, and r is in degrees Celsius per second.

[0085] Understandably, modules located deep within the cabinet with higher thermal resistance will experience a significantly higher rate of temperature rise at the same joule power than modules near the front vents. Continuing with the previous example, if Req is taken as 0.8 degrees Celsius per watt and τ as 60 seconds, then r is approximately 0.43 degrees Celsius per second.

[0086] Preferably, the temperature rise rate is calculated one by one for each unbalanced module in this manner to obtain a list of temperature rise rates corresponding to the unbalanced module numbers. The safe operating temperature rise limit represents the maximum allowable temperature rise rate per unit time of the module during the ramp-up phase, which comes from the module's manufacturer's specifications and is commonly found in the range of 0.5 degrees Celsius per second.

[0087] It should be noted that for modules with large heat dissipation deviation indicators, their current heat dissipation conditions have deviated from the calibration environment during factory testing, and directly using the factory limit value would overestimate their load-bearing capacity.

[0088] In one embodiment, based on the heat dissipation deviation index of the corresponding unbalanced module in the heat dissipation difference feature file, the safe operating temperature rise limit is linearly reduced according to the absolute value of the deviation index. The reduction amount is in the range of 0 to 0.15 degrees Celsius per second. After deducting the reduction amount, a safe temperature rise benchmark rlim adapted to the current heat dissipation conditions is obtained.

[0089] For example, if the heat dissipation deviation index is too large, the adjustment is taken as 0.12, and the original 0.5 degrees Celsius per second is corrected to rlim of 0.38 degrees Celsius per second. Furthermore, the temperature rise exceeding the limit reflects the degree to which the temperature rise rate caused by the distribution current exceeds the limit of the safe temperature rise reference.

[0090] Specifically, for each imbalanced module, Δr = r - rlim is calculated as the temperature rise exceeding the limit. Using the example, r = 0.43, rlim = 0.38, then Δr = 0.05 degrees Celsius per second.

[0091] Understandably, the absolute value of the temperature rise exceeding the limit depends on the differences in the physical parameters of each module, making direct horizontal comparison difficult.

[0092] Preferably, the degree of temperature rise exceeding the limit is divided by the safe temperature rise benchmark to obtain the degree of exceedance coefficient s = Δr / rlim, which is a dimensionless number. Following the example, s is approximately 0.13. If Δr is negative, s will naturally be negative. Furthermore, the probability of local overheating reflects the likelihood that the unbalanced module will undergo thermal derating or protective current limiting under the distributed current.

[0093] In one embodiment, the probability of local overheating is defined as the likelihood that the module will be subject to thermal derating protection while bearing the current distribution current. This probability is obtained by mapping the over-limit coefficient to a range of zero to one.

[0094] Specifically, a probability mapping based on the logistic function is used to map the excess degree coefficient s through p=1 / (1+e -k×(s-s0)The probability of overheating is converted into a value, where p represents the probability of local overheating, e represents the natural constant, k represents the slope parameter, and s0 represents the center offset parameter. The slope parameter k and the center offset parameter s0 are calibrated using statistics from previous thermal derating events. The calibration principle is as follows: samples of thermal derating protection occurring in historical operating data are collected, the actual protection triggering frequency corresponding to different over-limit coefficients is statistically analyzed, and the parameters of the logistic curve are fitted using maximum likelihood estimation to minimize the error between the mapped probability and the actual triggering frequency. Common calibration results show k = 10 and s0 = 0.05. This parameter combination results in a probability of approximately 0.5 when the over-limit coefficient is 0.05, and an increase in probability of approximately 0.2 to 0.3 for every 0.1 increase in the over-limit coefficient, consistent with the gradual triggering characteristics of thermal derating protection. Using the example, after mapping, p = 0.13 is approximately 0.69, indicating that the imbalance module has a relatively high probability of overheating during the ramp-up phase.

[0095] It should be noted that for unbalanced modules where s is negative or close to zero, the logistic function naturally maps them to a probability value close to zero.

[0096] For example, when s=-0.05, p is approximately 0.27; when s=0, p is approximately 0.38. The module number, temperature rise rate, over-limit coefficient, and local overheat probability of each unbalanced module are written into the local overheat probability file in rows to obtain the local overheat probability of each module with current imbalance risk, which can be used as a criterion for subsequent power redistribution.

[0097] S107. Compare the probability of local overheating with a preset probability threshold, obtain the remaining available output margin of the modules whose probability of local overheating is lower than the preset probability threshold, reallocate the output power according to the remaining available output margin, and determine the target allocation current of each module in the power ramp-up stage according to the voltage and output power of each module.

[0098] The local overheating probability of each module is obtained from the local overheating probability file and compared with a preset probability threshold. Modules with probabilities higher than the threshold are marked as restricted modules, and modules with probabilities lower than or equal to the threshold are included in the safe module set. For each module in the safe module set, the remaining usable output margin is obtained by multiplying the temperature rise space by the heat capacity time constant and then dividing by the product of the equivalent heat dissipation thermal resistance and the ramp-up phase duration. The total power that the set can carry is obtained by summing the remaining usable output margins of each module in the safe module set. The total power command for the ramp-up phase is received. If the total power command exceeds the total power that the set can carry, it is truncated to the total power that the set can carry. The obtained total power is divided according to the proportion of the remaining usable output margin of each module to obtain the redistributed output power. The redistributed output power is divided by the bus voltage sampling value of each safe module during the power ramp-up phase to obtain the target allocation current of the module. For the restricted modules, the target allocation current is set to zero to obtain the target allocation current of each module during the power ramp-up phase.

[0099] When the photovoltaic energy storage and charging system enters the next charging period and starts power ramp-up, the local overheat probability of each parallel module is read from the local overheat probability file as the core basis for subsequent allocation criteria.

[0100] It should be noted that the local overheating probability reflects the likelihood that the module will be subject to thermal derating protection when undertaking the current distribution. A preset probability threshold is used to classify power-bearing eligibility.

[0101] In one possible implementation, the probability threshold is set to 0.5.

[0102] Specifically, modules with a local overheating probability higher than 0.5 are marked as restricted modules, while modules with a local overheating probability lower than or equal to 0.5 are classified into the safe module set.

[0103] It is understandable that the restricted modules do not participate in power sharing during this period, while the members of the safe module set share the redistributed output power according to their respective carrying capacities.

[0104] Preferably, for nighttime charging periods or periods with high scheduling priority, the probability threshold can be tightened to 0.4 to provide a greater overheat avoidance margin. Nighttime charging periods refer to the charging operations performed by the charging station during off-peak electricity price periods, typically from 10 PM to 6 AM the following day. During this period, the charging power duration is longer and users are less sensitive to charging time. Periods with high scheduling priority refer to periods requiring priority response to emergency charging needs or grid dispatch instructions. During these periods, the charging task completion rate must be guaranteed. The charging work periods mentioned above typically last for more than 2 hours, longer than the 0.5 to 1 hour work period of regular charging. Therefore, stricter overheat probability control is needed to avoid the risk of heat accumulation due to prolonged operation. The work period is the duration of a complete charging task, the scheduling cycle is a repetitive cycle composed of multiple work periods, and the power ramp-up phase is the first 5 to 10 minutes of a single work period. The overheat avoidance margin ΔPm is defined as the difference between the current power level and the critical power that triggers overheat protection, in kilowatts. Its value is equal to the remaining available output margin after correction by a safety factor of 0.8 for the available power space. Furthermore, the remaining available output margin represents the maximum continuous power that each module in the safety module set is allowed to handle during the power ramp-up phase. The remaining available output margin is calculated based on four quantities: temperature rise margin ΔT, thermal capacity time constant τ, equivalent thermal resistance Req, and ramp-up phase duration t. The temperature rise margin ΔT represents the temperature difference between the module's current temperature and its overheat protection threshold, in degrees Celsius. It is obtained by subtracting the real-time temperature reading Tc from the module's built-in temperature sensor from the module's factory-calibrated overheat protection threshold Tth, i.e., ΔT = Tth - Tc. Different modules have different overheat protection thresholds due to differences in power device type and heat dissipation design. IGBT modules typically have a threshold of 85 degrees Celsius, while silicon carbide modules can reach 110 degrees Celsius. When the ambient temperature is 25 degrees Celsius and the initial module temperature is 45 degrees Celsius, the temperature rise margin for an IGBT module is 40 degrees Celsius, and for a silicon carbide module, it is 65 degrees Celsius. The thermal capacity time constant τ represents the time required for the module temperature to rise to 63.2% of its steady-state temperature rise, expressed in seconds. Its value is equal to the product of the module's thermal capacity C and its thermal resistance R, i.e., τ = C × R. The thermal capacity C is provided by the module manufacturer in the product datasheet; the thermal capacity time constant for typical power modules is 40 to 80 seconds, and 60 seconds is used in this embodiment. The equivalent thermal resistance Req represents the steady-state temperature rise caused by each watt of power dissipated by the module, expressed in degrees Celsius per watt. Its value comprehensively considers the module's own thermal resistance, the contact thermal resistance between the module and the heat sink, and the convection thermal resistance of the heat sink. For the front module, the airflow temperature is low and the flow rate is high, so the equivalent thermal resistance is taken as 0.5 degrees Celsius per watt. For the middle module, the airflow temperature increases and the flow rate decreases, so the equivalent thermal resistance is taken as 0.8 degrees Celsius per watt. The ramp-up phase duration t is determined based on the charging protocol handshake time and current slope limit, typically ranging from 300 to 600 seconds.

[0105] In one embodiment, the remaining available output margin Pavl is given by the following relationship: ΔT represents the allowable temperature rise of the module before entering the next working period, in degrees Celsius; τ represents the thermal capacity time constant of the module, in seconds; Req represents the equivalent thermal resistance of the module, in degrees Celsius per watt; and t represents the duration of the power ramp-up phase issued by the scheduling plan, in seconds.

[0106] For example, for a safety module near the front ventilation opening of the cabinet, with an ambient temperature rise ΔT of 40 degrees Celsius, a thermal time constant τ of 60 seconds, an equivalent thermal resistance Req of 0.5 degrees Celsius per watt, and a peak duration t of 30 seconds, the allowable peak power consumption can be calculated as Pavl = ΔT × τ / (Req × t) to be approximately 160 watts. For a safety module in the middle of the cabinet, affected by the heat accumulation of the preceding stage, with ΔT of 25 degrees Celsius, τ of 60 seconds, Req of 0.8 degrees Celsius per watt, and t of 30 seconds, the Pavl is approximately 62.5 watts.

[0107] Understandably, modules with larger temperature rise margins and larger heat capacity time constants have higher remaining usable output margins, while modules with higher thermal resistance or faster temperature rise have lower remaining usable output margins.

[0108] Specifically, the remaining available output margins of all modules within the safety module set are summed one by one to obtain the total power Psum that the set can carry. Using the previous example of two modules, assuming that the safety module set contains only these two modules, then Psum is approximately 222.5 watts. Further, the scheduling plan issues a total power command Pcmd for the climb phase, serving as the output target for this period. Pcmd is compared with Psum: if Pcmd does not exceed Psum, the original value of Pcmd is retained; if Pcmd exceeds Psum, Pcmd is truncated to Psum, resulting in the corrected total power value Pset.

[0109] For example, when Pcmd is 300 watts and Psum is 222.5 watts, Pset is set to 222.5 watts; when Pcmd is 180 watts, Pset is set to 180 watts. After Pset is determined, if Pset is less than Psum, Pset is proportionally divided according to the proportion of the remaining available output margin of each safety module in Psum; if Pset is equal to Psum, each module directly uses its remaining available output margin as the redistributed output power. Using the example, the remaining margin of the front module is 160 watts, and the remaining margin of the middle module is 62.5 watts. When Pset is 222.5 watts, each module directly uses its original margin value; when Pset is 180 watts, the proportion of the front module is 160÷222.5≈0.72 and the proportion of the middle module is 62.5÷222.5≈0.28. The front module receives 180×0.72≈129.6 watts and the middle module receives 180×0.28≈50.4 watts, thus obtaining the redistributed output power of each module.

[0110] It should be noted that the target distribution current is determined jointly by the redistribution output power and the bus voltage. The bus voltage is the DC terminal voltage shared by all parallel modules, obtained from the voltage sensor on the DC bus during the power ramp-up phase. To avoid current jitter caused by instantaneous ripple, the bus voltage is taken as a steady-state value at the beginning of each scheduling cycle, denoted as Ubus. In one embodiment, for each safety module, its redistribution output power is divided by the bus voltage at the corresponding moment, and the target distribution current of that module during the power ramp-up phase is obtained by Itar = Pre / Ubus, where Itar represents the target distribution current, Pre represents the redistribution output power of the module, and Ubus represents the bus voltage at the same moment. Following the example, when Ubus is 800 volts, the target distribution current of the front module is approximately 0.2 amps, and the target distribution current of the middle module is approximately 0.078 amps.

[0111] Preferably, the bus voltage sampling value can be taken as a steady-state value at the beginning of each scheduling cycle to avoid current jitter caused by instantaneous ripple. For the aforementioned restricted module, its target allocation current during the power ramp-up phase is directly set to zero, indicating that the restricted module does not respond to power commands during this period, remains idle, and cools down naturally.

[0112] It is understandable that the target allocation current is written into the target allocation file according to the module number and the cabinet partition to which it belongs. This serves as the current command to drive each energy storage converter during the power ramp-up phase. It matches the uneven temperature accumulated by each module in the cabinet during the interval period, thus preventing the module with the worst heat dissipation condition from reaching the thermal derating threshold too early and being forced to reduce output.

[0113] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for scheduling and controlling multi-time-period charging modes in a photovoltaic-storage-charging system, characterized in that, The method includes: quantifying the heat dissipation differences between modules based on their installation positions and cooling duct structures within the cabinet of the photovoltaic energy storage and charging system; determining the temperature locations of each module's substrate to be monitored based on these heat dissipation differences, collecting substrate temperatures, calculating the rate of change of each module's substrate temperature over time, and determining the cooling rate of each module; receiving the interval between adjacent charging periods given by the scheduling plan, and combining the cooling rate of each module with the current substrate temperature to infer the estimated temperature of each module's substrate at the end of the interval; comparing the estimated substrate temperature of each module with the upper limit of the safe operating temperature, and analyzing the allowable temperature rise space of each module before entering the next working period. The process involves obtaining the temperature rise margin for each module; allocating the expected total current during the photovoltaic output ramp-up phase according to the temperature rise margin of each module, determining whether the allocated current for each module matches its temperature rise margin, and identifying the risk of current imbalance; estimating the temperature rise rate of each module with current imbalance risk when bearing the allocated current based on the heat dissipation thermal resistance of the module with current imbalance risk and its allocated current, and determining the probability of local overheating of each module with current imbalance risk based on the degree to which the temperature rise rate exceeds the safe operating temperature rise limit; comparing the probability of local overheating with a preset probability threshold, obtaining the remaining usable output margin of modules with a local overheating probability lower than the preset probability threshold, and redistributing the output power based on the remaining usable output margin.

2. The method for scheduling and controlling multi-time-period charging modes of a photovoltaic energy storage and charging system according to claim 1, characterized in that, The method for quantifying the heat dissipation differences between modules based on the installation position and cooling airflow structure of each energy storage module in the rack of the photovoltaic energy storage and charging system includes: obtaining the installation position coordinates of each module in the rack, dividing it into rows and columns and recording the distance of each module to the nearest air vent and the upstream and downstream positions of the airflow to form a spatial position file; obtaining the initial thermal resistance value of each module's substrate to the environment based on the spatial position file, superimposing the thermal coupling resistance value estimated by the heat source density and spacing of adjacent modules to obtain the equivalent heat dissipation thermal resistance and forming a thermal resistance distribution table; calculating the arithmetic mean of the thermal resistance of each module in the thermal resistance distribution table, using the difference between the thermal resistance of each module and the average value as the heat dissipation deviation index, and combining the airflow path to identify fast-cooling modules and slow-cooling modules, and incorporating them into the heat dissipation difference feature file.

3. The method for scheduling and controlling multi-time-period charging modes of a photovoltaic energy storage and charging system according to claim 2, characterized in that, The process of determining the temperature locations to be monitored for each module substrate based on the heat dissipation difference characteristics, collecting substrate temperatures, calculating the rate of change of each module substrate temperature over time, and determining the cooling rate of each module includes: determining measurement points by placing points at the projection positions directly above the power devices of each module and at the edge positions of the heat dissipation substrate according to the fast and slow cooling module labels in the heat dissipation difference characteristic file; attaching thermistors at the measurement points to collect substrate temperatures at a preset sampling rate to obtain a base temperature sequence; after sliding window mean filtering, taking the absolute value of the ratio of the temperature difference between adjacent time windows to the time window duration as the cooling rate of each module and archiving it according to the module number.

4. The method for scheduling and controlling multi-time-period charging modes of a photovoltaic energy storage and charging system according to claim 1, characterized in that, The method of receiving the interval duration of adjacent charging periods given by the scheduling plan, and inferring the estimated temperature of each module's substrate at the end of the adjacent charging period interval by combining the cooling rate of each module and the current substrate temperature, includes: dividing the interval duration into a time step sequence according to a preset step size, collecting the ambient temperature inside the cabinet as a reference value for the final cooling temperature; establishing a time step-by-time temperature decay recursive relationship based on the cooling rate of each module and the difference between the current substrate temperature and the reference value, accumulating the temperature drop along the time step sequence to obtain the temperature drop trajectory, and archiving the end temperature value corresponding to the end of the interval to form an estimated temperature file.

5. The method for scheduling and controlling multi-time-period charging modes of a photovoltaic energy storage and charging system according to claim 1, characterized in that, The process of comparing the estimated substrate temperature of each module with the upper limit of the safe operating temperature and analyzing the allowable temperature rise space of each module before entering the next working period to obtain the temperature rise margin of each module includes: obtaining the estimated temperature of the substrate of each module from the estimated temperature file; adjusting the initial upper limit of the safe operating temperature of the module at the factory according to the heat dissipation deviation index to obtain the upper limit value; taking the difference between the estimated temperature and the upper limit value as the temperature rise space of each module; adding a lenient label to the module whose temperature rise space is higher than the preset warning threshold, and adding an urgent label to the module whose temperature rise space is lower than the preset warning threshold, and importing them into the temperature rise margin file.

6. The method for scheduling and controlling multi-time-period charging modes of a photovoltaic energy storage and charging system according to claim 1, characterized in that, The process of allocating the expected total current for the photovoltaic output ramp-up phase according to the temperature rise margin of each module, determining whether the allocated current for each module matches its temperature rise margin, and identifying current imbalance risks includes: using the proportion of each module's temperature rise space in the total as the allocation weight, proportionally dividing the expected total current for the ramp-up phase issued by the scheduling plan according to the allocation weight to obtain the allocated current for each module; using the product of the square of the allocated current and the equivalent heat dissipation thermal resistance of the module as the temperature rise estimate, subtracting the corresponding temperature rise space from the temperature rise estimate to obtain the matching difference; adding an imbalance label to modules with a matching difference higher than a preset imbalance threshold and marking them as having current imbalance risk, adding a normal label to modules with a matching difference lower than the imbalance threshold, and importing the module number and corresponding label into the imbalance risk file.

7. The method for scheduling and controlling multi-time-period charging modes of a photovoltaic energy storage and charging system according to claim 1, characterized in that, The step of estimating the temperature rise rate of each module at risk of current imbalance when bearing the distributed current, based on the thermal resistance of the module with current imbalance risk and its distributed current, includes: obtaining the module number with imbalance identifier and the corresponding distributed current; retrieving the equivalent thermal resistance from the thermal resistance distribution table for each imbalanced module and reading the on-resistance and thermal capacity time constant of the power device; using the product of the square of the distributed current and the on-resistance as the Joule power; and dividing the product of the Joule power and the equivalent thermal resistance by the thermal capacity time constant to obtain the temperature rise rate of each imbalanced module.

8. The method for scheduling and controlling multi-time-period charging modes of a photovoltaic energy storage and charging system according to claim 1, characterized in that, The determination of the local overheating probability of each module with current imbalance risk based on the degree to which the temperature rise rate exceeds the safe operating temperature rise limit includes: retrieving the safe operating temperature rise limit for each imbalance module, adjusting the safe operating temperature rise limit downward according to the heat dissipation deviation index to obtain a safe temperature rise benchmark; subtracting the safe temperature rise benchmark from the temperature rise rate to obtain the temperature rise exceeding the limit, dividing the temperature rise exceeding the limit by the safe temperature rise benchmark to obtain an exceeding degree coefficient; and performing a nonlinear transformation on the exceeding degree coefficient using a probability mapping based on a logistic function to obtain a probability value mapped to between 0 and 1, which serves as the local overheating probability of each module with current imbalance risk.

9. A method for scheduling and controlling multi-time-period charging modes in a photovoltaic energy storage and charging system according to claim 1, characterized in that, The step of comparing the probability of local overheating with a preset probability threshold to obtain the remaining usable output margin of modules whose probability of local overheating is lower than the preset probability threshold includes: comparing the probability of local overheating of each module with the preset probability threshold, marking modules with a probability higher than the preset probability threshold as restricted modules, and classifying modules with a probability lower than or equal to the preset probability threshold into a set of safe modules; for each module in the set of safe modules, multiplying the temperature rise space by the heat capacity time constant and then dividing by the product of the equivalent heat dissipation thermal resistance and the climb-up stage duration to obtain the remaining usable output margin.

10. A method for scheduling and controlling multi-time-period charging modes in a photovoltaic energy storage and charging system according to claim 9, characterized in that, After obtaining the remaining available output margin, the process includes: summing the remaining available output margins of each module in the set of safety modules to obtain the total power that the set can carry; receiving the total power command during the ramp-up phase, whereby the total power command is truncated to the total power that the set can carry when it exceeds the total power that the set can carry; proportionally dividing the obtained total power according to the proportion of the remaining available output margin of each module to obtain the redistributed output power; dividing the redistributed output power by the bus voltage sampling value of each safety module during the power ramp-up phase to obtain the target allocation current of the module; and setting the target allocation current to zero for the restricted modules.

Images