Energy storage cabinet temperature control method based on cloud platform

Abstract

The application discloses a kind of energy storage cabinet temperature control methods based on cloud platform, it is related to energy storage cabinet temperature control technical field, the collaborative architecture of this application through cloud platform and edge computing, realize multilayer temperature control optimization;Based on the conduction modeling of dynamic graph neural network, accurately identify the implicit thermal coupling between adjacent cabinet, trigger precooling in the early stage of thermal runaway, inhibit chain reaction;Fusion barometric pressure, photovoltaic fluctuation and other multi-source data, dynamically adjust air cooling / liquid cooling strategy, avoid the heat dissipation failure caused by single environmental mutation;Through priority communication slice and the lightweight control logic of edge computing unit, ensure the low-delay execution of key instruction, break through the delay bottleneck of traditional cloud centralized control.

Description

Technical Field

[0001] This invention relates to the field of energy storage cabinet temperature control technology, and in particular to an energy storage cabinet temperature control method based on a cloud platform. Background Technology

[0002] With the rapid deployment of new energy power plants and industrial and commercial energy storage scenarios, energy storage systems are evolving towards high density and clustering. Containerized energy storage units composed of multiple parallel cabinets have become one of the mainstream solutions due to their flexible deployment and convenient expansion. However, there is a significant thermal coupling effect between densely arranged battery cabinets in a limited space: the heat released by thermal runaway of a single cabinet can be quickly conducted to adjacent cabinets through radiation, convection and other means, forming a chain reaction. In such scenarios, the traditional single-cabinet independent temperature control mode is difficult to cope with cross-device thermal interference. Although new temperature control technologies such as liquid cooling and phase change cooling have improved local heat dissipation efficiency, cluster-level thermal safety risks continue to exist.

[0003] Most current solutions incorporate digital twin and distributed edge computing technologies. For example, a digital twin-based thermal simulation system predicts the temperature distribution between cabinets through 3D modeling and links liquid cooling valves to adjust the flow rate (such as the intelligent liquid cooling solution launched by Sungrow Power in 2023). Some manufacturers use distributed edge nodes to deploy lightweight AI models to analyze the temperature trend inside the cabinet in real time and trigger early warnings. However, these solutions still have two major drawbacks: First, thermal coupling modeling relies on static physical parameters (such as fixed spacing and constant wind speed) and cannot dynamically capture the implicit correlation changes caused by sudden environmental changes (such as strong crosswinds changing the heat convection path).

[0004] To address the aforementioned issues, some solutions employ federated learning frameworks to optimize cross-cabinet data collaboration or introduce topology analysis algorithms to identify high-risk transmission paths. However, such solutions are still limited by hardware compatibility and real-time requirements. Communication delays between heterogeneous devices may lead to policy lags, and most edge controllers lack sufficient computing power to support complex topology calculations. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] This invention provides a cloud-based energy storage cabinet temperature control method to solve the problem that existing solutions rely on static models and single-cabinet control, which cannot predict heat conduction paths and have high response delays, leading to the risk of cluster-level thermal runaway.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] This invention provides a cloud-based method for temperature control of an energy storage cabinet, comprising:

[0009] Step S1: Obtain real-time operating data of each battery cabinet in the energy storage system, including battery temperature, charging and discharging power and environmental parameters, and upload the operating data to the cloud platform;

[0010] Step S2, the cloud platform generates a heat conduction prediction model based on the operating data, the model including the thermal coupling relationship between each battery cabinet and environmental disturbance factors;

[0011] Step S3: Based on the heat conduction prediction model, dynamically generate a temperature control strategy for each battery cabinet. The temperature control strategy includes liquid cooling flow distribution instructions, air cooling start-up and shutdown sequence, and phase change material activation threshold.

[0012] Step S4: The temperature control strategy is sent to the edge computing unit of each energy storage cabinet, and the edge computing unit executes the corresponding liquid cooling pump speed adjustment, fan control and valve opening and closing operations.

[0013] As a preferred embodiment of the cloud platform-based energy storage cabinet temperature control method of the present invention, in step S1, the operating data further includes photovoltaic string current fluctuation data; the cloud platform predicts the charging and discharging power change trend of each battery cabinet within a future preset time period based on the current fluctuation data, and inputs the trend into the heat conduction prediction model;

[0014] The preset time period is dynamically set based on the movement speed of the photovoltaic array's shadow, specifically:

[0015] When the detected string current fluctuation rate exceeds 10% / second, the time period is set to the next 5 minutes.

[0016] When the fluctuation rate is less than 5% / second, the time period is set to the next 15 minutes.

[0017] As a preferred embodiment of the cloud platform-based energy storage cabinet temperature control method of the present invention, in step S2, the heat conduction prediction model is constructed by a graph neural network, wherein the nodes represent the real-time temperature status of the battery cabinet, and the edge weights represent the heat radiation efficiency and convection influence coefficient between adjacent cabinets; the heat radiation efficiency and convection influence coefficient are dynamically updated according to the cabinet spacing, layout orientation and real-time wind direction.

[0018] As a preferred embodiment of the cloud-based energy storage cabinet temperature control method described in this invention, the dynamic update logic for the thermal radiation efficiency and convection influence coefficient includes:

[0019] Real-time wind direction data is obtained by wind speed sensors deployed on the outer surface of the energy storage cabinet;

[0020] Calculate the heat convection correction factor based on the relative orientation between the cabinets;

[0021] When strong crosswinds are detected, the weight of the convection impact coefficient is increased.

[0022] As a preferred embodiment of the cloud platform-based energy storage cabinet temperature control method of the present invention, step S2 further includes:

[0023] The cloud platform aggregates historical thermal runaway event data from multi-regional energy storage systems through a federated learning framework and continuously updates the thermal coupling parameters in the thermal conduction prediction model.

[0024] The historical thermal runaway event data includes heat conduction paths, the effectiveness of intervention measures, and the rate of temperature decay.

[0025] As a preferred embodiment of the cloud platform-based energy storage cabinet temperature control method of the present invention, in step S3, the temperature control strategy further includes an air pressure adaptive parameter, and the cloud platform adjusts the operation mode of the air-cooled system based on the air pressure value in real-time meteorological data.

[0026] When the current air pressure is detected to be lower than the preset threshold, the air-cooling mode is switched to intermittent pulse operation, and the reference flow rate of the liquid cooling system is increased simultaneously.

[0027] As a preferred embodiment of the cloud platform-based energy storage cabinet temperature control method of the present invention, in step S3, when the heat conduction prediction model determines that a certain battery cabinet has a risk of abnormal temperature, the cloud platform dynamically adjusts the temperature warning threshold of the adjacent cabinets and sends a pre-cooling command to the edge computing unit of the adjacent cabinets.

[0028] The pre-cooling instructions include starting the liquid cooling system in advance and limiting the maximum speed of the air cooling system.

[0029] As a preferred embodiment of the cloud platform-based energy storage cabinet temperature control method of the present invention, in step S3, when the current air pressure is detected to be lower than a preset threshold, the air-cooling mode is switched to intermittent pulse operation, the steps include:

[0030] Air pressure values ​​collected from the edge unit With threshold After comparison, the following conditions are met:

[0031] ,

[0032] This triggers the pulse mode and defines the air pressure deviation:

[0033] ,

[0034] Based on this, the on-time and off-time of the fan within each pulse cycle are dynamically calculated:

[0035] ,

[0036] ,

[0037] The complete cycle can be obtained from this:

[0038] ,

[0039] And duty cycle:

[0040] ,

[0041] Edge units are configured according to the above timing sequence at intervals of... Starts in seconds The fan stops in seconds. Seconds, continuously looping.

[0042] in, This represents the real-time atmospheric pressure value collected by the edge unit. This indicates the preset air pressure switching threshold. This represents the difference between the threshold and the measured pressure. This indicates the duration of the fan operation within a single pulse cycle. A scaling factor representing the duration of operation. The nonlinear modulation index represents the on-time duration. This indicates the duration of the fan shutdown within a single pulse cycle. A scaling factor representing the duration of shutdown. The nonlinear modulation index represents the off-time. Indicates the duration of the complete pulse cycle. This indicates the proportion of the cycle that the fan is in operation.

[0043] As a preferred embodiment of the cloud platform-based energy storage cabinet temperature control method described in this invention, in step S3, when the heat conduction prediction model determines the number as... When a battery cabinet poses a risk of abnormal temperature, the cloud platform aggregates all its adjacent cabinets. Each cabinet Perform the following actions:

[0044] Risk indicators detected from model output ,scope If the risk threshold is exceeded, the thermal coupling coefficient will be used as the basis for determining the risk level. Calculate and update the temperature warning thresholds for adjacent cabinets:

[0045] ,

[0046] in, Indicates cabinet The original temperature warning threshold, in °C. This represents the threshold adjustment scaling factor, which is dimensionless. Indicates cabinet and The thermal coupling coefficient between them is dimensionless. Indicates cabinet Temperature anomaly risk index, dimensionless. This indicates the updated warning threshold, in °C.

[0047] Simultaneously, the cloud platform generates a pre-cooling command and sends it to the edge unit. The command includes starting liquid cooling in advance and increasing the flow rate according to the risk level, as well as limiting the maximum speed of the air-cooled system.

[0048] ,

[0049] in, Indicates cabinet The reference liquid cooling flow rate, in L / min. This represents the flow rate increase coefficient, which is dimensionless. Indicates cabinet The nominal maximum fan speed, in rpm. This represents the air-cooled speed limiting factor, dimensionless, range. , This indicates the adjusted liquid cooling flow rate, in L / min. This indicates the adjusted maximum air-cooled speed, in rpm.

[0050] As a preferred embodiment of the cloud platform-based energy storage cabinet temperature control method of the present invention, in step S4, the communication between the cloud platform and the edge computing unit adopts a multi-level priority channel allocation mechanism; for control commands with a temperature anomaly risk level higher than a preset threshold, the transmission is carried out through independent communication slices, and the response delay does not exceed 200ms.

[0051] The independent communication slice transmission is achieved in the following way:

[0052] In 5G NR networks, a dedicated ultra-reliable low-latency communication (URLLC) channel is allocated for temperature anomaly commands;

[0053] A dual-link redundant transmission mechanism is adopted, with the primary link being a cellular network and the backup link being a LoRa self-organizing network.

[0054] The beneficial effects of this invention are: this invention achieves multi-level temperature control optimization through a collaborative architecture of cloud platform and edge computing;

[0055] Based on the conduction modeling of dynamic graph neural networks, the implicit thermal coupling between adjacent cabinets is accurately identified, and pre-cooling is triggered in the early stage of thermal runaway to suppress the chain reaction.

[0056] By integrating multi-source data such as air pressure and photovoltaic fluctuations, the air-cooling / liquid-cooling strategy is dynamically adjusted to avoid heat dissipation failure caused by a single environmental change.

[0057] By prioritizing communication slices and using lightweight control logic from edge computing units, we ensure low-latency execution of critical instructions, overcoming the latency bottleneck of traditional cloud-based centralized control. Attached Figure Description

[0058] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This is a flowchart illustrating a cloud-based energy storage cabinet temperature control method in Example 1. Detailed Implementation

[0060] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0061] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0062] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0063] Example 1, referring to Figure 1 This embodiment provides a cloud-based method for controlling the temperature of an energy storage cabinet, including the following steps:

[0064] Step S1: Obtain real-time operating data of each battery cabinet in the energy storage system, including battery temperature, charging and discharging power and environmental parameters, and upload the operating data to the cloud platform;

[0065] In step S1, the operating data also includes photovoltaic string current fluctuation data; the cloud platform predicts the charging and discharging power change trend of each battery cabinet within a preset time period based on the current fluctuation data, and inputs the trend into the heat conduction prediction model;

[0066] The preset time period is dynamically set based on the movement speed of the photovoltaic array's shadow, specifically as follows:

[0067] When the detected string current fluctuation rate exceeds 10% / second, the time period is set to the next 5 minutes.

[0068] When the fluctuation rate is less than 5% / second, the time period is set to the next 15 minutes; this can be set according to the actual rate and needs.

[0069] Step S2: The cloud platform generates a heat conduction prediction model based on the operating data. The model includes the thermal coupling relationship between each battery cabinet and environmental disturbance factors.

[0070] In step S2: the heat conduction prediction model is constructed through a graph neural network, where nodes represent the real-time temperature status of the battery cabinet, and edge weights represent the heat radiation efficiency and convection influence coefficient between adjacent cabinets; the heat radiation efficiency and convection influence coefficient are dynamically updated according to the cabinet spacing, layout orientation and real-time wind direction.

[0071] The dynamic update logic for thermal radiation efficiency and convection influence coefficient includes:

[0072] Real-time wind direction data is obtained by wind speed sensors deployed on the outer surface of the energy storage cabinet;

[0073] Calculate the heat convection correction factor based on the relative orientation between the cabinets;

[0074] When strong crosswinds are detected, the weight of the convective impact coefficient is increased.

[0075] Step S2 also includes:

[0076] The cloud platform aggregates historical thermal runaway event data from multi-regional energy storage systems through a federated learning framework, and continuously updates the thermal coupling parameters in the heat conduction prediction model.

[0077] Historical thermal runaway event data includes heat conduction pathways, effectiveness of intervention measures, and temperature decay rates;

[0078] Step S3: Based on the heat conduction prediction model, dynamically generate temperature control strategies for each battery cabinet. The temperature control strategies include liquid cooling flow distribution instructions, air cooling start-up and shutdown sequence, and phase change material activation threshold.

[0079] In step S3, the temperature control strategy also includes an adaptive air pressure parameter, and the cloud platform adjusts the operating mode of the air-cooling system based on the air pressure value in the real-time meteorological data.

[0080] When the current air pressure is detected to be lower than the preset threshold, the air cooling mode is switched to intermittent pulse operation, and the reference flow rate of the liquid cooling system is increased simultaneously.

[0081] In step S3, when the heat conduction prediction model determines that a certain battery cabinet has a risk of abnormal temperature, the cloud platform dynamically adjusts the temperature warning threshold of the adjacent cabinets and sends a pre-cooling command to the edge computing unit of the adjacent cabinets.

[0082] Pre-cooling instructions include starting the liquid cooling system in advance and limiting the maximum speed of the air cooling system;

[0083] In step S3, when the current air pressure is detected to be lower than a preset threshold, the air-cooling mode is switched to intermittent pulse operation. The steps include:

[0084] Air pressure values ​​collected from the edge unit With threshold After comparison, the following conditions are met:

[0085] ,

[0086] This triggers the pulse mode and defines the air pressure deviation:

[0087] ,

[0088] Based on this, the on-time and off-time of the fan within each pulse cycle are dynamically calculated:

[0089] ,

[0090] ,

[0091] The complete cycle can be obtained from this:

[0092] ,

[0093] And duty cycle:

[0094] ,

[0095] Edge units are configured according to the above timing sequence at intervals of... Starts in seconds The fan stops in seconds. Seconds, continuously looping.

[0096] in, This represents the real-time atmospheric pressure value collected by the edge unit. This indicates the preset air pressure switching threshold. This represents the difference between the threshold and the measured pressure. This indicates the duration of the fan operation within a single pulse cycle. A scaling factor representing the duration of operation. The nonlinear modulation index represents the on-time duration. This indicates the duration of the fan shutdown within a single pulse cycle. A scaling factor representing the duration of shutdown. The nonlinear modulation index represents the off-time. Indicates the duration of the complete pulse cycle. This indicates the proportion of the cycle that the fan is on;

[0097] Specifically, this pulse-type operation strategy dynamically adjusts the fan on and off time based on air pressure deviation, without manual intervention. It can quickly extend the fan on time in low air pressure environments to compensate for the decrease in heat dissipation efficiency. At the same time, it automatically reduces the fan duty cycle when the air pressure recovers to save energy and reduce wear. By calculating and issuing control in real time at the edge unit, it ensures that the command latency is much lower than the single cycle length, making it suitable for energy storage cabinet scenarios that are sensitive to temperature changes and response delays. In addition, the scaling coefficient and exponential factor can be flexibly adjusted to adapt to different cabinet layouts and environmental conditions, and it has good versatility and scalability.

[0098] In step S3, when the heat conduction prediction model determines the number as... When a battery cabinet poses a risk of abnormal temperature, the cloud platform aggregates all its adjacent cabinets. Each cabinet Perform the following actions:

[0099] Risk indicators detected from model output ,scope If the risk threshold is exceeded, the thermal coupling coefficient will be used as the basis for determining the risk level. Calculate and update the temperature warning thresholds for adjacent cabinets:

[0100] ,

[0101] in, Indicates cabinet The original temperature warning threshold, in °C. This represents the threshold adjustment scaling factor, which is dimensionless. Indicates cabinet and The thermal coupling coefficient between them is dimensionless. Indicates cabinet Temperature anomaly risk index, dimensionless. This indicates the updated warning threshold, in °C.

[0102] Simultaneously, the cloud platform generates a pre-cooling command and sends it to the edge unit. The command includes starting liquid cooling in advance and increasing the flow rate according to the risk level, as well as limiting the maximum speed of the air-cooled system.

[0103] ,

[0104] in, Indicates cabinet The reference liquid cooling flow rate, in L / min. This represents the flow rate increase coefficient, which is dimensionless. Indicates cabinet The nominal maximum fan speed, in rpm. This represents the air-cooled speed limiting factor, dimensionless, range. , This indicates the adjusted liquid cooling flow rate, in L / min. This indicates the adjusted maximum air-cooled speed, in rpm;

[0105] Specifically, this method precisely adjusts the temperature warning threshold of adjacent cabinets by calculating the product of risk indicators and thermal coupling coefficients, so that abnormal risks can be detected at the earliest stage and trigger multi-level cooling responses. In the pre-cooling command, the liquid cooling system is started in advance and the flow rate is increased according to the risk intensity, while the maximum speed of air cooling is limited to prevent excessive impact, effectively balancing cooling efficiency and energy consumption. Compared with the traditional single-point alarm that only starts cooling, this strategy has foresight, adaptability and local linkage, significantly improving the overall temperature control reliability and response speed of the energy storage cabinet, and can be flexibly optimized by scaling coefficients to adapt to different cabinet layouts and load conditions.

[0106] Step S4: The temperature control strategy is sent to the edge computing unit of each energy storage cabinet, and the edge computing unit executes the corresponding liquid cooling pump speed adjustment, fan control and valve opening and closing operations.

[0107] In step S4, the communication between the cloud platform and the edge computing unit adopts a multi-level priority channel allocation mechanism; for control commands with a temperature anomaly risk level higher than the preset threshold, they are transmitted through independent communication slices, and the response delay does not exceed 200ms.

[0108] Independent communication slice transmission is achieved in the following ways:

[0109] In 5G NR networks, a dedicated ultra-reliable low-latency communication (URLLC) channel is allocated for temperature anomaly commands;

[0110] A dual-link redundant transmission mechanism is adopted, with the primary link being a cellular network and the backup link being a LoRa self-organizing network.

[0111] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for temperature control of an energy storage cabinet based on a cloud platform, characterized in that, include, Step S1: Obtain real-time operating data of each battery cabinet in the energy storage system, including battery temperature, charging and discharging power and environmental parameters, and upload the operating data to the cloud platform; Step S2, the cloud platform generates a heat conduction prediction model based on the operating data, the model including the thermal coupling relationship between each battery cabinet and environmental disturbance factors; Step S3: Based on the heat conduction prediction model, dynamically generate a temperature control strategy for each battery cabinet. The temperature control strategy includes liquid cooling flow distribution instructions, air cooling start-up and shutdown sequence, and phase change material activation threshold. Step S4: The temperature control strategy is sent to the edge computing unit of each energy storage cabinet, and the edge computing unit executes the corresponding liquid cooling pump speed adjustment, fan control and valve opening and closing operations.

2. The method for temperature control of an energy storage cabinet based on a cloud platform as described in claim 1, characterized in that, In step S1, the operating data also includes photovoltaic string current fluctuation data; the cloud platform predicts the charging and discharging power change trend of each battery cabinet within a future preset time period based on the current fluctuation data, and inputs the trend into the heat conduction prediction model; The preset time period is dynamically set based on the movement speed of the photovoltaic array's shadow, specifically: When the detected string current fluctuation rate exceeds 10% / second, the time period is set to the next 5 minutes. When the fluctuation rate is less than 5% / second, the time period is set to the next 15 minutes.

3. The method for temperature control of an energy storage cabinet based on a cloud platform as described in claim 1, characterized in that, In step S2, the heat conduction prediction model is constructed using a graph neural network, where nodes represent the real-time temperature status of the battery cabinet, and edge weights represent the heat radiation efficiency and convection influence coefficient between adjacent cabinets; the heat radiation efficiency and convection influence coefficient are dynamically updated based on the cabinet spacing, layout orientation, and real-time wind direction.

4. The method for temperature control of an energy storage cabinet based on a cloud platform as described in claim 3, characterized in that, The dynamic update logic for the thermal radiation efficiency and convection influence coefficient includes: Real-time wind direction data is obtained by wind speed sensors deployed on the outer surface of the energy storage cabinet; Calculate the heat convection correction factor based on the relative orientation between the cabinets; When strong crosswinds are detected, the weight of the convection impact coefficient is increased.

5. The method for temperature control of an energy storage cabinet based on a cloud platform as described in claim 4, characterized in that, Step S2 further includes: The cloud platform aggregates historical thermal runaway event data from multi-regional energy storage systems through a federated learning framework and continuously updates the thermal coupling parameters in the thermal conduction prediction model. The historical thermal runaway event data includes heat conduction paths, the effectiveness of intervention measures, and the rate of temperature decay.

6. The method for temperature control of an energy storage cabinet based on a cloud platform as described in claim 1, characterized in that, In step S3, the temperature control strategy also includes an adaptive air pressure parameter, and the cloud platform adjusts the operating mode of the air-cooling system based on the air pressure value in real-time meteorological data. When the current air pressure is detected to be lower than the preset threshold, the air-cooling mode is switched to intermittent pulse operation, and the reference flow rate of the liquid cooling system is increased simultaneously.

7. The method for temperature control of an energy storage cabinet based on a cloud platform as described in claim 6, characterized in that, In step S3, when the heat conduction prediction model determines that a certain battery cabinet has a risk of abnormal temperature, the cloud platform dynamically adjusts the temperature warning threshold of the adjacent cabinets and sends a pre-cooling command to the edge computing unit of the adjacent cabinets. The pre-cooling instructions include starting the liquid cooling system in advance and limiting the maximum speed of the air cooling system.

8. The method for temperature control of an energy storage cabinet based on a cloud platform as described in claim 7, characterized in that, In step S3, when the current air pressure is detected to be lower than a preset threshold, the air-cooling mode is switched to intermittent pulse operation. The steps include: Air pressure values ​​collected from the edge unit With threshold After comparison, the following conditions are met: ， This triggers the pulse mode and defines the air pressure deviation: ， Based on this, the on-time and off-time of the fan within each pulse cycle are dynamically calculated: ， ， The complete cycle can be obtained from this: ， And duty cycle: ， Edge units are configured according to the above timing sequence at intervals of... Starts in seconds The fan stops in seconds. Seconds, continuously looping. in, This represents the real-time atmospheric pressure value collected by the edge unit. This indicates the preset air pressure switching threshold. This represents the difference between the threshold and the measured pressure. This indicates the duration of the fan operation within a single pulse cycle. A scaling factor representing the duration of operation. The nonlinear modulation index represents the on-time duration. This indicates the duration of the fan shutdown within a single pulse cycle. A scaling factor representing the duration of shutdown. The nonlinear modulation index represents the off-time. Indicates the duration of the complete pulse cycle. This indicates the proportion of the cycle that the fan is in operation.

9. The method for temperature control of an energy storage cabinet based on a cloud platform as described in claim 8, characterized in that, In step S3, when the heat conduction prediction model determines the number as... When a battery cabinet poses a risk of abnormal temperature, the cloud platform aggregates all its adjacent cabinets. Each cabinet Perform the following actions: Risk indicators detected from model output ,scope If the risk threshold is exceeded, the thermal coupling coefficient will be used as the basis for determining the risk level. Calculate and update the temperature warning thresholds for adjacent cabinets: ， in, Indicates cabinet The original temperature warning threshold, in °C. This represents the threshold adjustment scaling factor, which is dimensionless. Indicates cabinet and The thermal coupling coefficient between them is dimensionless. Indicates cabinet Temperature anomaly risk index, dimensionless. This indicates the updated warning threshold, in °C. Simultaneously, the cloud platform generates a pre-cooling command and sends it to the edge unit. The command includes starting liquid cooling in advance and increasing the flow rate according to the risk level, as well as limiting the maximum speed of the air-cooled system. ， in, Indicates cabinet The reference liquid cooling flow rate, in L / min. This represents the flow rate increase coefficient, which is dimensionless. Indicates cabinet The nominal maximum fan speed, in rpm. This represents the air-cooled speed limiting factor, dimensionless, range. , This indicates the adjusted liquid cooling flow rate, in L / min. This indicates the adjusted maximum air-cooled speed, in rpm.

10. The method for temperature control of an energy storage cabinet based on a cloud platform as described in claim 1, characterized in that, In step S4, the communication between the cloud platform and the edge computing unit adopts a multi-level priority channel allocation mechanism; for control commands with a temperature anomaly risk level higher than a preset threshold, they are transmitted through independent communication slices, and the response delay does not exceed 200ms. The independent communication slice transmission is achieved in the following way: In 5G NR networks, a dedicated ultra-reliable low-latency communication (URLLC) channel is allocated for temperature anomaly commands; A dual-link redundant transmission mechanism is adopted, with the primary link being a cellular network and the backup link being a LoRa self-organizing network.

Images