Quick starting method and device for onboard power module

Abstract

The invention relates to the technical field of quick starting of power supplies, and discloses a quick starting method and device for an onboard power module, and the method comprises the steps: carrying out the pre-charging processing of a distributed power supply, deploying a quick starting strategy, optimizing the pre-charging state in real time through a dynamic reference slope algorithm and a self-adaptive current amplitude limiting algorithm, generating monitoring information, and carrying out the quick starting of the onboard power module. Time-frequency domain joint analysis is carried out on monitoring information, a lightweight CNN model is input to generate power supply load characteristics, a pre-charging state is reconstructed based on the load characteristics, a starting strategy is adjusted, and when a starting instruction is monitored, the power supply is started according to an optimization strategy, and the method can improve the starting stability and efficiency of the power supply, ensure quick response and load balance of the system, and improve the starting efficiency of the power supply. The method adapts to a dynamic environment, and solves the problem that in the prior art, fixedly set quick start cannot dynamically adapt to the real-time condition of the battery.

Description

Technical Field

[0001] The present invention relates to the technical field of rapid power startup, and in particular to a rapid startup method and device for a board-mounted power module. Background Art

[0002] With the increasing demand for power supply stability, fast response, and high efficiency in electronic devices, traditional power supply startup methods have shown certain shortcomings in certain application scenarios. This is especially true in distributed power supply systems, where power supply startup stability and timeliness are key factors affecting system performance. Distributed power supplies typically include multiple power supply units, which have different load behaviors and response characteristics. During the startup process, each power supply unit needs to work in coordination to avoid problems such as overload and instability. Traditional power supply startup methods often suffer from slow response, especially in systems with multiple parallel power supplies. Coordination of the startup process is particularly complex. Factors such as the power supply pre-charging process and current fluctuations during startup can affect the stability of the entire system. Currently, the pre-charging state and startup strategy of most power modules are based on fixed settings and do not account for the dynamic changes in power supply load behavior. This can lead to unstable startup of the power supply system in actual operation. In most existing technologies, the pre-charging state and startup strategy of the power supply lack real-time optimization mechanisms, making it impossible to adjust them promptly according to power supply load changes, thus affecting the startup performance of the power supply module. Summary of the Invention

[0003] The object of the present invention is to provide a method and device for quickly starting a board-mounted power module, aiming to solve the problem in the prior art that the fixed-setting quick start cannot dynamically adapt to the real-time status of the battery.

[0004] The present invention is implemented as follows: In a first aspect, the present invention provides a method for quickly starting a board-mounted power supply module, comprising: Perform pre-charging processing on the distributed power supply pre-installed on the circuit board, and deploy corresponding fast startup strategies for the distributed power supply in the specified pre-charging state; The pre-charging state is optimized in real time by a dynamic reference ramp algorithm and an adaptive current limiting algorithm, and power supply pre-charging monitoring information is generated; Performing a joint time-frequency domain analysis on the power pre-charge monitoring information, and substituting the analysis results into a lightweight CNN model deployed at the edge to generate power load behavior characteristics; Reconstructing a specified form of the pre-charge state of the distributed power supply according to the power load behavior characteristics, and synchronously adjusting the fast startup strategy; When a power startup instruction is monitored, the distributed power supply is quickly started according to the fast startup strategy.

[0005] In a second aspect, the present invention provides a rapid startup device for an onboard power module, which is used to implement the rapid startup method for an onboard power module described in any one of the first aspects, comprising: A pre-start module is used to pre-charge the distributed power supply pre-installed on the circuit board and deploy a corresponding fast start strategy for the distributed power supply in a specified pre-charge state; A real-time optimization module, configured to optimize the pre-charging state in real time by using a dynamic reference ramp algorithm and an adaptive current limiting algorithm, and to generate power pre-charging monitoring information; A feature monitoring module is used to perform a joint time-frequency domain analysis on the power pre-charging monitoring information and substitute the analysis results into a lightweight CNN model deployed at the edge to generate power load behavior characteristics; a layout reconstruction module, configured to reconstruct a specified form of a pre-charge state of the distributed power supply according to the power load behavior characteristics, and synchronously adjust the fast startup strategy; The fast startup module is used to quickly start the distributed power supply according to the fast startup strategy when a power startup instruction is monitored.

[0006] The present invention provides a method for quickly starting a board-mounted power supply module, which has the following beneficial effects: The present invention performs pre-charging on distributed power supplies and deploys a fast startup strategy. It optimizes the pre-charging state in real time through a dynamic reference ramp algorithm and an adaptive current limiting algorithm, generates monitoring information, performs a joint time-frequency domain analysis on the monitoring information, and inputs a lightweight CNN model to generate power supply load characteristics. Based on the load characteristics, the pre-charging state is reconstructed and the startup strategy is adjusted. When a startup instruction is monitored, the power supply is started according to the optimization strategy. This method can improve the stability and efficiency of power supply startup, ensure fast system response, load balance, and adapt to dynamic environments, solving the problem in the prior art that the fixed fast startup cannot dynamically adapt to the real-time status of the battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 This is a schematic diagram of the steps of a quick startup method for an onboard power module provided by an embodiment of the present invention; Figure 2 The diagram is a structural diagram of a quick start device for a board-mounted power module provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0008] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0009] The implementation of the present invention is described in detail below with reference to specific embodiments.

[0010] Reference Figure 1 、 Figure 2 As shown, a preferred embodiment of the present invention is provided.

[0011] In a first aspect, the present invention provides a method for quickly starting a board-mounted power supply module, comprising: S1: pre-charge the distributed power supplies pre-installed on the circuit board and deploy corresponding fast startup strategies for the distributed power supplies in the specified pre-charge state; S2: Optimizing the pre-charge state in real time through a dynamic reference ramp algorithm and an adaptive current limiting algorithm, while generating power supply pre-charge monitoring information; S3: Performing a joint time-frequency domain analysis on the power pre-charging monitoring information, and substituting the analysis results into a lightweight CNN model deployed at the edge to generate power load behavior characteristics; S4: reconstructing the designated form of the pre-charge state of the distributed power supply according to the power load behavior characteristics, and synchronously adjusting the fast startup strategy; S5: When a power supply startup instruction is monitored, the distributed power supply is quickly started according to the fast startup strategy.

[0012] Specifically, in step S1 of the embodiment provided by the present invention, the distributed power supply is pre-charged. The pre-charging process makes basic preparations for the startup of the power module. Before startup, the power module is gradually put into working state through pre-charging to ensure smooth operation at startup. Through proper pre-charging, excessive impact or damage to the power supply unit caused by direct power startup can be avoided. Pre-charging can reduce the surge current during startup and prevent damage to the power supply unit. The pre-charging process can effectively reduce the instantaneous loss of the power supply unit at startup and increase the service life of the system. Through a stable pre-charging process, it is ensured that each power supply unit can work stably at a state close to the rated value, avoiding unstable behavior at startup. The pre-charged power supply unit can adapt to load changes more quickly when entering the normal working state, thereby improving the overall system efficiency.

[0013] More specifically, a corresponding fast start-up strategy is deployed for distributed power supplies in a specified form of pre-charging state. The pre-charging state of each distributed power supply unit is different, and an adaptive start-up strategy needs to be deployed for it. By adjusting the start-up strategy according to the pre-charging state and load behavior characteristics of each power supply unit, the startup process can be made more accurate and efficient. By dynamically adjusting the fast start-up strategy according to the power load behavior characteristics, it can respond in time when the actual load changes, optimize the startup process, and avoid startup failure or instability due to load changes. According to the pre-charging state and requirements of different power supply units, the fast start-up strategy can ensure that the power module is put into work quickly and reduce startup delays. By reasonably configuring the startup strategy, the various power supply units can work together, avoiding startup conflicts between different power supply units and improving the startup efficiency of the entire system. By deploying a corresponding fast start-up strategy for each power supply unit, unnecessary resource waste during the startup process can be avoided and energy utilization can be improved.

[0014] Specifically, in step S2 of the embodiment provided by the present invention, a dynamic reference ramp algorithm is applied to perform pre-charging optimization. The dynamic reference ramp algorithm sets a reference ramp to enable the power supply unit to achieve a smooth current transition during the startup process. The algorithm can adjust the ramp curve according to the real-time load and power supply status to ensure that the current change of the power supply unit is not too drastic, thereby avoiding damage to the power supply unit caused by current shock. By dynamically adjusting the reference ramp, the current change of the power supply unit can be accurately controlled to avoid excessive fluctuations, thereby improving the predictability and stability of the startup process. Through the smooth ramp algorithm, the current surge at the startup of the power supply unit is reduced to avoid excessive pressure on the power supply unit and improve the reliability of the system. The dynamic reference ramp algorithm can adjust the pre-charging curve according to real-time changes, so that the power supply unit enters the working state more smoothly, thereby improving the smoothness of the overall system startup. The dynamic reference ramp can adapt to load changes in real time and flexibly adjust the current curve to ensure smooth startup under different load conditions.

[0015] More specifically, an adaptive current limiting algorithm is implemented to limit current. The adaptive current limiting algorithm can limit current according to actual current demand and dynamically adjust the upper limit of current during the pre-charging process to avoid overload of the power supply unit. The algorithm can adjust the current upper limit in real time according to the pre-charging status and load demand to ensure that the current of the power supply unit does not exceed the maximum carrying capacity, thereby protecting the power supply unit. Through adaptive current limiting, the current upper limit can be adjusted in real time to prevent the power supply unit from being damaged due to excessive current during the startup process, ensuring the safety of the power supply unit during the startup process. By avoiding the impact of excessive current, the power supply unit can operate within a safe range, reducing losses caused by overload, thereby extending the service life of the system. The adaptive current limiting algorithm can dynamically adjust the current according to actual demand, which not only protects the power supply unit, but also effectively avoids energy waste.

[0016] More specifically, the pre-charging state is optimized in real time. Through the combination of dynamic reference ramp and adaptive current limiting algorithm, the pre-charging state of the power supply unit can be optimized according to real-time monitoring data. The real-time adjustment of the pre-charging state can dynamically adapt to the needs of the power supply unit according to factors such as load changes and temperature changes, ensuring its optimal state at startup. Real-time optimization of the pre-charging state helps to balance the load of the power supply unit during the startup process, making the startup faster and more stable. Real-time optimization can adjust the pre-charging process in time according to different load conditions, so that the power supply unit can start efficiently and stably under various loads. By optimizing the pre-charging state of the power supply in real time, the system can quickly adjust the working state of the power supply unit when the load or environment changes, thereby improving the startup response speed. The system can respond to external environmental changes in real time, automatically adjust the working mode of the power supply unit, and improve the adaptability of the overall system.

[0017] More specifically, generating power pre-charging monitoring information is an important part of real-time monitoring and data recording of the entire power module. By recording the current, voltage, temperature and other parameters of each power unit, it can provide valuable data support for subsequent maintenance, debugging and optimization. The monitoring information can provide real-time feedback on the operating status of the power unit, help detect possible abnormal conditions, and take corresponding measures to make adjustments to avoid system problems. By real-time monitoring and generating power pre-charging information, potential problems in the system can be discovered in time, and early fault warnings can be provided, which helps to improve the system's maintainability and fault recovery capabilities. The monitoring data provides a comprehensive view of the system operation, which can help detect potential faults or performance degradations in advance, and improve fault diagnosis and prediction capabilities. Real-time monitoring data provides strong support for subsequent system optimization and debugging, and helps adjust the system according to the actual operating status to ensure that it is always in the best working condition.

[0018] Specifically, in step S3 of the embodiment provided by the present invention, the power pre-charging monitoring information is jointly analyzed in the time and frequency domains. The pre-charging state of the power supply will be affected by many factors. Relying solely on the time domain or frequency domain may not be able to fully describe the power supply behavior. The time-frequency domain joint analysis can simultaneously capture the time-varying characteristics and frequency characteristics of the power supply signal, so that the understanding of the power supply load behavior is more comprehensive. During the power supply pre-charging process, parameters such as current and voltage may exhibit more complex fluctuation behaviors when they change instantaneously. The time-frequency joint analysis can accurately capture these instantaneous characteristics and help extract features with physical significance. The time-frequency domain joint analysis can more accurately identify key features in the power supply pre-charging process, such as mutation points, frequency changes, noise, etc., thereby helping to improve the accuracy of the power supply load behavior analysis. The time-frequency domain analysis can effectively remove noise interference, maintain the effective information of the signal, and help improve the system's robustness and adaptability to interference.

[0019] More specifically, the time-frequency domain analysis results are substituted into the lightweight CNN model. The lightweight convolutional neural network (CNN) can effectively process the large amount of feature data obtained by the joint analysis of the time-frequency domain and extract the key patterns in the power load behavior. Compared with the traditional deep learning model, the lightweight CNN is more suitable for edge device deployment, with lower computing requirements and fast response speed. By deploying the CNN model on the edge device, real-time data processing can be performed locally in the power system, reducing the delay and bandwidth pressure caused by data transmission, and improving the overall system response speed. The lightweight CNN can efficiently identify and extract the characteristics of the power load, such as load fluctuations, the frequency and amplitude of load changes, etc., so as to accurately classify and predict the power load behavior. The design of the lightweight CNN model enables it to run on resource-constrained edge devices, achieve efficient computing, avoid transmitting large amounts of data to the cloud or central server, improve the system's processing capability and real-time performance, and deploy the CNN model to the edge computing node so that the system can quickly respond and process changes in the power load in real time, avoiding the slow response problem caused by network delays.

[0020] More specifically, power load behavior characteristics are generated by substituting power pre-charge monitoring information into a lightweight CNN model. The resulting power load behavior characteristics provide operational data support for subsequent power performance evaluation, fault diagnosis, load prediction, etc. The generated load behavior characteristics can help identify the regularity and abnormal fluctuations of the power load, and provide a data basis for power load regulation, fault warning, maintenance optimization, etc. By analyzing the power load behavior characteristics, the changing trend of the power load can be predicted, and measures can be taken in advance to deal with possible problems such as overload and power fluctuation, thereby improving the predictability of the power system. Based on the generated load behavior characteristics, intelligent power management can be realized, load scheduling can be optimized, energy utilization efficiency can be improved, and the service life of power equipment can be extended. By real-time monitoring and generating power load behavior characteristics, power load anomalies can be identified in advance, fault warning mechanisms can be activated, and faults can be detected and repaired in a timely manner to avoid system shutdown or damage.

[0021] More specifically, the load behavior characteristics are associated and optimized with other systems. The load behavior characteristics are not only helpful for power management and scheduling, but can also be associated with other systems (such as temperature management, environmental monitoring, etc.). The overall operation of the power system is optimized through multi-dimensional information fusion. The power load behavior characteristics are combined with the data of other monitoring systems to achieve cross-system collaborative optimization and further improve the stability and efficiency of the system. By fusing the load behavior characteristics with the data of other systems, the overall performance of the system can be improved, and multi-angle optimization can be performed to ensure that the power system, environmental monitoring system, etc. can work together to achieve efficient and stable power management. The comprehensive utilization of load behavior characteristics can provide a basis for the global management strategy of the power system, such as optimizing the pre-charging process, adjusting power output, controlling environmental impact, etc., thereby improving the efficiency of overall energy management.

[0022] More specifically, edge deployment and feedback mechanism. Edge deployment can carry out data processing, analysis and decision-making processes as close to the power supply equipment as possible, reduce transmission delays, and improve overall response speed. Through local calculation and local feedback, the system can make adjustment decisions faster and improve real-time performance. Through real-time feedback on load behavior characteristics, the system can adjust the working mode of the power supply and perform optimization adjustments, such as load balancing and current regulation. Edge deployment can significantly reduce the delay in data transmission and processing, and enhance the system's ability to respond quickly under changing conditions. Through edge computing and real-time feedback mechanism, the system can automatically adjust according to load behavior characteristics, enhancing the adaptability and flexibility of power load management.

[0023] It is understandable that through the technical steps of joint analysis in the time and frequency domains, lightweight CNN model processing, and load behavior feature generation, the changing characteristics of the power load can be comprehensively and accurately captured and optimized. The time and frequency domain analysis provides multi-dimensional signal characteristics, the lightweight CNN model ensures efficient real-time processing, and the load behavior feature generation provides strong data support for the intelligent management, fault warning, and load prediction of the power system. Ultimately, these steps can greatly improve the stability, response speed, and intelligence level of the power system.

[0024] Specifically, in step S4 of the embodiment provided by the present invention, the specified form of the pre-charging state is reconstructed according to the power supply load behavior characteristics. The power supply load behavior characteristics (such as load fluctuations, frequency changes, overload conditions, etc.) can directly reflect the actual working state of the power supply. Reconstructing the pre-charging state according to these characteristics can ensure that the power supply is in the optimal pre-charging condition, thereby improving the efficiency and stability of the system. The power supply load behavior will change over time, and a fixed pre-charging state may not be able to cope with changing load requirements. By dynamically reconstructing the pre-charging state according to the load behavior characteristics, the system can adapt to different working environments and load requirements in real time, avoiding resource waste or power supply damage caused by improper pre-charging. Through real-time monitoring and analysis of the power supply load behavior, the pre-charging requirements of the power supply under different load conditions can be accurately identified and optimized and adjusted, thereby improving the accuracy and efficiency of the pre-charging process. Dynamic adjustment of the pre-charging state can prevent energy waste caused by excessive pre-charging of the power supply or startup difficulties caused by insufficient pre-charging, ensuring that the power supply can be put into use under optimal conditions.

[0025] More specifically, the fast startup strategy is adjusted synchronously. The core of the fast startup strategy is to make adjustments based on the current state of the power supply (such as pre-charging state, load demand, etc.) to ensure that the power supply can work stably in the shortest time. The load behavior characteristics provide real-time power supply status information. Adjusting the fast startup strategy based on this information can ensure that the power supply avoids unnecessary delays or overloads during the startup process. The load behavior characteristics of the power supply directly affect the smoothness of the startup process. According to the load characteristics, timely adjustment of the startup strategy can avoid current overload, slow startup and other problems when the power supply starts, and improve the response speed and stability of the system. By adjusting the startup strategy according to the load behavior characteristics of the power supply, a faster and smoother startup can be achieved, avoiding startup delays or instability caused by excessive or mismatched loads. Dynamic adjustment of the startup strategy can effectively utilize the pre-charging state and load information of the power supply, while ensuring stable startup of the power supply, optimizing energy consumption and reducing unnecessary power waste.

[0026] More specifically, real-time adjustments are made and feedback mechanisms are implemented. The load behavior characteristics in the power supply system are constantly changing, especially in complex distributed power supply networks. Through real-time monitoring and feedback mechanisms, the adjustment of the pre-charging state and the startup strategy can be made more flexible and precise. The effectiveness of the fast startup strategy depends on the synchronous changes in the pre-charging state and the power load. By implementing the feedback mechanism, the change information of the load behavior can be obtained in real time during the power supply operation, so as to dynamically adjust the strategy to ensure that the power supply system can operate efficiently and stably under various working conditions. Through real-time feedback and dynamic adjustment, the system can adapt to the different challenges brought about by load changes, ensure the flexibility and adaptability of power load management, and further improve the performance of the power supply. Real-time feedback adjustment can help to promptly discover and correct potential problems in the power supply system, reduce the impact of overload, underload and other problems on power supply operation, thereby improving the reliability and stability of the system.

[0027] More specifically, optimize the collaborative work between distributed power sources. In a distributed power supply system, each power source is affected by different loads, environments, operating conditions and other factors. By optimizing the collaborative work between each power source, the overall balance and efficiency of the power supply system can be ensured. The reconstructed pre-charge state and fast start-up strategy should consider the collaborative work of multiple power sources to avoid load imbalance or resource waste. The collaborative work strategy of the distributed power supply system requires information sharing and collaborative scheduling between power sources. By optimizing the startup, load distribution, power output, etc. of each power source, it can ensure that the system can respond quickly and maintain stable operation when facing complex load requirements. By optimizing the collaborative work between each distributed power source, the system can efficiently dispatch power resources, improve the coordination between power sources, and enhance the overall load management capability. The collaborative work of multiple power sources can ensure load balancing between power sources, avoid overloading or underutilization of certain power sources, thereby optimizing the allocation of power resources and improving the overall system efficiency.

[0028] More specifically, long-term performance monitoring and adjustment are implemented. The working state of the power supply system will change with changes in time, load and other factors. Especially in distributed power supply systems, the interactions between power supplies and environmental factors may change continuously. Therefore, implementing long-term performance monitoring and adjusting the pre-charging state and startup strategy based on the monitoring data is the key to ensuring the long-term stable operation of the system. Through long-term monitoring, more power load behavior data can be obtained, which helps to optimize the long-term operation strategy of the power supply and make accurate load predictions to avoid the impact of emergencies on the power supply system. Through the implementation of long-term performance monitoring and adjustment, the system can make long-term responses to power load changes to ensure that the power supply system always remains efficient and stable during long-term operation. Long-term monitoring can provide the power supply system with more accurate prediction models and adjustment strategies, so that the power supply system can better adapt to future load changes and reduce sudden failures and resource waste.

[0029] It can be understood that by reconstructing the pre-charging state of distributed power sources and synchronously adjusting the technical steps of the fast startup strategy, precise control and optimization of the power supply system can be achieved, and the pre-charging state and startup strategy can be dynamically adjusted according to the power load behavior characteristics. This not only improves the startup efficiency of the power supply and the response speed of the system, but also optimizes the coordinated work of multiple power sources to ensure long-term stable operation of the system. In addition, by implementing real-time feedback and long-term monitoring mechanisms, the system can continuously optimize power load management and resource allocation, and improve the performance and reliability of the overall power management system.

[0030] Specifically, in step S5 of the embodiment provided by the present invention, a power startup instruction is monitored. Monitoring of the power startup instruction is a prerequisite for rapid response in the system. After monitoring the power startup instruction, the system can immediately start the relevant program to enable the power supply. Real-time monitoring of the instruction ensures that the power supply can respond immediately when needed to avoid delays. After monitoring the power startup instruction, the startup information can be obtained in the first time to ensure that the system can execute subsequent startup operations as soon as possible, avoiding unnecessary time waste due to waiting or delays. After real-time monitoring of the startup instruction, the system can respond quickly and trigger a quick startup process to ensure that the startup operation is not affected by any delay factors, thereby improving the response speed and efficiency of the system. The accuracy and timeliness of instruction monitoring ensure the smooth progress of the startup process and reduce system instability that may be caused by erroneous instructions or delayed responses.

[0031] More specifically, the power supply startup conditions are evaluated according to the quick startup strategy. The quick startup strategy is not just a simple startup, but it evaluates the current state of the power supply (such as pre-charge state, power supply load, etc.) to ensure that the startup conditions meet the preset optimal state. Through strategy evaluation, the system can confirm that the startup operation will not cause overload, power waste and other problems. The quick startup strategy adjusts the startup method according to the load and health status of different power supplies. The startup conditions and loads of each distributed power supply may be different. Different startup strategies are adopted after evaluation to avoid system overload. By evaluating the startup conditions, the system can automatically adjust the startup strategy according to the status of the power supply to ensure that the startup process meets the optimal working conditions and avoid failure or inefficiency caused by mismatched conditions. Evaluating the startup conditions makes the startup of each power supply more accurate and reduces the risk of system instability or damage caused by improper startup.

[0032] More specifically, a quick startup operation is performed. After confirming that the power supply status meets the quick startup conditions, the system starts to perform the quick startup operation, which includes operations such as power switching, load distribution, and power calibration. This step ensures that the power supply can be stably started in a short time and is ready for use. The quick startup strategy provides preset startup methods for different loads, states, and requirements. These methods may include sudden load management, phased startup, and other operations. By accurately executing the preset plan, the startup process can be smoother and energy use can be optimized. Performing a quick startup operation can significantly shorten the time from the power supply receiving the startup instruction to the full commissioning, improve the system response speed, and ensure that the power supply can meet the load requirements as soon as possible. Through optimized startup operations, not only the quick startup of the power supply is guaranteed, but also the efficiency of the power supply at startup can be improved, reducing energy waste.

[0033] More specifically, power output is automatically adjusted based on load demand. After startup, power output needs to be adjusted based on real-time load. If the load demand is too high, the power system should increase output power; if the load demand is low, the output power should be reduced to achieve efficient energy utilization. A key feature of distributed power sources is their flexible load adaptability. By automatically adjusting output based on load demand, the system can ensure efficient operation under any operating conditions. Dynamically adjusting power output can prevent certain power supplies from overloading or unstable operation due to unbalanced loads, thereby improving the overall stability and efficiency of the system. Automatically adjusting power output based on real-time load demand can reduce unnecessary power consumption, improve the overall system's energy efficiency, and reduce operating costs.

[0034] More specifically, the startup status is monitored and adjusted in real time. After the power supply is started, the system should continue to monitor the operating status of the power supply to ensure that it operates in accordance with the fast startup strategy and adjust any abnormalities that may occur. After the power supply is quickly started, abnormal conditions may still occur due to load changes or environmental factors. Real-time adjustment can avoid these problems. Through continuous monitoring and adjustment, the system can promptly detect and handle potential startup failures or performance problems to ensure that the power supply system always operates stably. Through real-time monitoring and adjustment, it can quickly respond to various emergencies and avoid system failures due to startup abnormalities or load imbalance.

[0035] More specifically, a feedback mechanism and optimization strategy are implemented. Through the feedback mechanism during the startup process, power supply operation data can be collected in real time and fed back to the system for subsequent adjustment and optimization. The feedback mechanism can identify optimization space that may appear during the startup process and continuously adjust the strategy. Through continuous feedback, the power supply system can adjust the strategy according to historical data and real-time operating conditions to improve the long-term efficiency and stability of the system. The feedback mechanism allows the system to continuously optimize the fast startup strategy according to actual operating conditions, making each startup operation more efficient. Through optimization strategy and feedback adjustment, the system can adaptively adjust the operation mode according to different working environments and load conditions to ensure the long-term high efficiency of the power supply system.

[0036] It can be understood that by taking the technical steps of quickly starting the distributed power supply according to the quick start strategy when a power start instruction is monitored, the system can achieve efficient and stable power start-up and load management. Each step has a clear execution reason and technical effect. From real-time monitoring to load adaptation, to feedback optimization, it can ensure that the power supply system can maintain efficient and stable operation when facing different working conditions. Ultimately, by optimizing startup time, reducing energy consumption, and improving system responsiveness, this solution can significantly improve the overall performance and reliability of the distributed power supply system.

[0037] The present invention provides a method for quickly starting a board-mounted power supply module, which has the following beneficial effects: The present invention performs pre-charging on distributed power supplies and deploys a fast startup strategy. It optimizes the pre-charging state in real time through a dynamic reference ramp algorithm and an adaptive current limiting algorithm, generates monitoring information, performs a joint time-frequency domain analysis on the monitoring information, and inputs a lightweight CNN model to generate power supply load characteristics. Based on the load characteristics, the pre-charging state is reconstructed and the startup strategy is adjusted. When a startup instruction is monitored, the power supply is started according to the optimization strategy. This method can improve the stability and efficiency of power supply startup, ensure fast system response, load balance, and adapt to dynamic environments, solving the problem in the prior art that the fixed fast startup cannot dynamically adapt to the real-time status of the battery.

[0038] Preferably, the steps of pre-charging the distributed power supply pre-installed on the circuit board and deploying a corresponding fast startup strategy for the distributed power supply in a specified pre-charge state include: S11: acquiring power supply structure information of a distributed power supply pre-set on a circuit board, so as to allocate unit connection marks for describing electrical connection relationships to each of a plurality of parallel-connected basic power supply units in the distributed power supply; S12: performing a quick startup gradient analysis on each of the basic power supply units based on the power supply configuration information of the distributed power supply, so as to assign a startup gradient mark to each of the basic power supply units; S13: specifying a pre-charging specification for each basic power supply unit according to the startup gradient flag assigned to each basic power supply unit, so as to perform a pre-charging process on each basic power supply unit in accordance with the specified pre-charging specification through a pre-charging circuit, so that each basic power supply unit is in a pre-charging state in accordance with the specified pre-charging specification, thereby placing the distributed power supply in a specified pre-charging state; S14: Based on the startup gradient mark and unit connection mark assigned to each of the basic power supply units, as well as the pre-charging specifications of each of the basic power supply units, a priority control architecture analysis for power supply rapid startup is performed on each of the basic power supply units to generate a rapid startup strategy for the distributed power supply in a specified form of pre-charging state.

[0039] Specifically, obtain the power supply structure information and assign unit connection marks. Before deploying the quick startup strategy, you first need to obtain the power supply structure information of the distributed power supply, including the type, connection method, layout, etc. of the power supply unit. This information is crucial for subsequent processing. Each basic power supply unit has multiple parallel units on the circuit board, and the electrical connection relationship between them must be clearly marked. Assigning unit connection marks is an important step to ensure that each power supply unit can be accurately interconnected and reasonably scheduled during the startup process. Assigning connection marks to each power supply unit ensures that the connection relationship between each power supply unit is clear during power startup and circuit scheduling, avoiding the risk of incorrect connection or connection confusion. The power supply structure information provides basic data for subsequent startup gradient analysis and pre-charge specification specification.

[0040] More specifically, a quick startup gradient analysis is performed based on the power supply structure information, and a startup gradient tag is assigned. Different power supply units may have different startup methods due to differences in type, load, structure, etc. By performing startup gradient analysis on the power supply units, appropriate startup sequences and intensities can be formulated for different units to avoid excessive load or current shock when parallel power supply units are started at the same time. The startup gradient tag helps to calibrate which power supply units should be started first and which power supply units can be added in subsequent gradient starts, thereby ensuring load balance and stability during system startup. Through startup gradient analysis, a reasonable startup sequence is assigned to each power supply unit to ensure system stability during startup and avoid current shock or startup failure. According to the startup requirements and gradient strategy of each power supply unit, the load shock caused by inconsistent startup of power supply units is reduced to avoid system instability.

[0041] More specifically, pre-charging specifications are assigned to each basic power supply unit, and pre-charging processing is performed. Each basic power supply unit of the distributed power supply needs to be prepared for startup through a pre-charging circuit. The pre-charging specifications are defined based on the power supply characteristics and startup gradient marks of each unit to ensure that each power supply unit can be in a stable state before startup. By specifying the pre-charging specifications, the initial voltage, current and other states of each power supply unit can be effectively controlled to avoid instability or damage during startup. Each power supply unit is started in a state that meets the pre-charging specifications, which can avoid adverse startup phenomena such as current overload, excessively high or too low voltage, and improve the stability of the system. Appropriate pre-charging specifications are formulated for each power supply unit to ensure that the power supply unit can quickly and reliably enter the startup state, thereby improving the reliability and startup efficiency of the system.

[0042] More specifically, based on the gradient mark, connection mark and pre-charge specifications, the priority control architecture is analyzed. Different power supply units may require different current, power and other conditions during the startup process. Through a comprehensive analysis of the startup gradient mark, connection mark and pre-charge specifications, the most appropriate priority control architecture can be designed so that the startup process of each power supply unit can be coordinated and consistent, thereby ensuring that the startup of the entire power supply system is more efficient. According to the connection relationship and startup gradient between the power supply units, the priority control architecture can help adjust the startup sequence of different power supply units to ensure that the power supply system starts smoothly and quickly. By analyzing and adjusting the startup priority of each power supply unit, startup conflicts between units are avoided, the coordination of the startup process is improved, and the startup process of each power supply unit is guaranteed to be smooth and efficient. After optimizing the control architecture, the power supply units can be started in the optimal order, thereby reducing startup delays and improving the response speed of the entire system.

[0043] More specifically, a quick startup strategy is generated. Through the analysis of the aforementioned steps, the final generated quick startup strategy will include information such as the startup sequence, pre-charge specifications, and startup gradient of each basic power supply unit. This strategy ensures that each power supply unit can be enabled in the best way when the power is started, and the system startup can be completed in the shortest time. This strategy is optimized based on the actual needs of different power supply units and the power structure. It can ensure that all operations in the startup process are in line with expectations, avoid excessive startup time or unnecessary waste of system resources. The generated startup strategy can ensure that each power supply unit is started in the optimal order, reduce startup time and avoid interference during the startup process. The optimized strategy can improve the startup efficiency of the overall system. Through precise strategy control, each power supply unit can be started at the appropriate time and in an appropriate manner, thereby improving the success rate and reliability of system startup.

[0044] It can be understood that through the above steps, the technical solution of pre-charging distributed power sources and deploying a fast startup strategy for them ensures the efficiency and stability of the power supply system during the startup process. From obtaining power supply structure information, analyzing startup gradients, allocating pre-charging specifications to generating a fast startup strategy, each step ensures coordination and efficient startup between power units while optimizing the overall performance of the power supply system to the greatest extent. The implementation of this solution can effectively reduce energy waste during startup, reduce system failure rate, and improve the system's response speed and long-term operation stability.

[0045] Preferably, the step of performing a quick startup gradient analysis on each of the basic power supply units based on the power supply configuration information of the distributed power supply to assign a startup gradient mark to each of the basic power supply units includes: S121: performing digital simulation of electrical parameter correlation, electrical connection topology, and thermal distribution for the distributed power supply based on the power supply configuration information, and using the digital simulation results as multiple constraints to perform adaptability evaluation of each basic power supply unit in the distributed power supply as a primary pre-charging target, so as to obtain adaptability characteristics of each basic power supply unit as a primary pre-charging target; S122: performing preliminary allocation of priority gradients and common gradients to the distributed power sources according to the adaptability characteristics of each of the basic power supply units as a primary pre-charging target, as an original gradient division form; S123: Performing a value assessment on the original gradient division form through a pre-built distributed power supply rapid startup knowledge base, iteratively optimizing the gradient allocation of the original gradient division form, and allocating startup gradient marks to each of the basic power supply units based on the optimized allocation form.

[0046] Specifically, based on the power supply structure information, digital simulation of electrical parameter correlation, electrical connection topology and thermal distribution is performed. In order to ensure a smooth startup process of the power supply unit, it is first necessary to simulate the electrical characteristics of each power supply unit (such as voltage, current, and power). This is to analyze the performance parameters of each power supply unit and their correlation with each other through the power supply structure information to determine the startup behavior of the power supply unit. The electrical connection relationship of the power supply unit determines how they work together at startup. Therefore, simulating the electrical connection topology structure helps to understand the interaction between different power supply units and their impact on the startup of the overall system. The thermal management of the power supply unit is crucial to the startup process, especially under high load. Under load conditions, uneven heat distribution may affect the performance of the power supply unit. Through digital simulation of heat distribution, potential thermal runaway areas can be identified before startup to avoid equipment damage or startup failure due to overheating. Through digital simulation, the electrical performance of the power supply unit under different operating conditions can be accurately predicted to ensure that each power supply unit does not encounter problems such as overload or voltage instability during startup. After simulating the electrical connection topology, the connection method of each power supply unit can be reasonably configured to reduce possible electrical shocks and improve system stability. Through thermal distribution analysis, hot spots can be identified in advance and optimized design can be performed to avoid equipment damage due to excessive temperature and ensure the long-term reliability of the system.

[0047] More specifically, the adaptability of each basic power supply unit as the main pre-charging object is evaluated. The adaptability of each basic power supply unit is evaluated based on the electrical parameters and thermal management analysis results. The purpose is to determine the performance of each power supply unit when it is a pre-charging object. Each power supply unit may perform differently under different pre-charging conditions and therefore needs to be evaluated separately. The adaptability evaluation not only focuses on the electrical characteristics of the power supply unit, but also needs to consider their reactions under heat, load and other conditions to ensure their stability and reliability during the pre-charging process. Through the adaptability evaluation, it is possible to accurately identify whether each power supply unit is suitable as the main object of pre-charging, avoiding unnecessary risks during the startup process. Through the selection after the adaptability evaluation, it can be ensured that the status of each power supply unit during the pre-charging process meets the startup requirements, thereby reducing the risk of failure.

[0048] More specifically, priority gradients and normal gradients are preliminarily allocated to form an original gradient division form. Different power supply units have large differences in the current and power required at startup. Therefore, the purpose of fast startup gradient allocation is to reasonably allocate startup gradients according to the characteristics of the power supply units to ensure that the power supply units are started in the best order. According to the adaptability characteristics obtained in the previous steps, the power supply units are divided into two categories: priority gradients and normal gradients. The power supply units with priority gradients can be started first, while the power supply units with normal gradients are started later. Such preliminary allocation helps to quickly optimize the startup process. Through preliminary gradient allocation, it can ensure that the power supply units with priority startup will not be subject to excessive interference, reduce the load impact during the startup process, and accelerate system startup through gradient allocation, avoiding startup delays caused by the simultaneous startup of all power supply units.

[0049] More specifically, the value of the original gradient division form is evaluated through the pre-built distributed power supply rapid startup knowledge base. The distributed power supply rapid startup knowledge base contains empirical data on the startup of different power supply units under various conditions. By evaluating the original gradient division form, it can be determined whether it conforms to the optimal startup strategy. Through the data in the knowledge base, the preliminary allocated gradient can be analyzed to evaluate whether it has practical feasibility, and necessary adjustments to the gradient allocation can be made based on the evaluation results. Through value evaluation, it is ensured that the original gradient division form can meet actual needs and avoid unreasonable startup strategies. By optimizing the gradient allocation through evaluation and feedback, the startup process is made more efficient and the startup of the power supply unit is more coordinated and stable.

[0050] More specifically, the original gradient division is iteratively optimized for gradient allocation. After preliminary evaluation, the accuracy and effectiveness of the allocation can be gradually improved through iterative optimization of the gradient allocation. Each iteration will be adjusted based on new feedback until the most suitable gradient allocation method is found. After each optimization, the gradient allocation will become more precise, ensuring that the power supply units are started in the best order and manner, minimizing the instability during the system startup process. Through iterative optimization, not only the startup success rate of the power supply units is improved, but also the startup time can be shortened, and the response speed of the overall system is improved. Gradually adjusting and optimizing the gradient allocation helps to ensure the stability of the system startup process and avoid problems such as current and voltage fluctuations.

[0051] More specifically, startup gradient tags are assigned to each basic power supply unit based on the optimized allocation form, and finally, startup gradient tags are assigned to each basic power supply unit based on the optimized gradient allocation form. These tags will directly affect the operation of the power supply units during the startup process, ensuring that they are started in the optimized order. Through precise startup gradient tag allocation, each power supply unit can be started at the appropriate time and in the appropriate order, thereby ensuring a fast and smooth startup of the system. The allocation of gradient tags can ensure that the power supply units work in sequence and coordination during the startup process, avoiding conflicts during the startup process. Through precise gradient tag allocation, each power supply unit can be started at the most appropriate time, further improving the startup reliability of the system.

[0052] It can be understood that through the above six steps, the distributed power supply is quickly analyzed for startup gradients and startup gradient marks are assigned, which can significantly improve the startup speed and efficiency while ensuring system stability. The execution of each step is based on the above analysis results, ensuring the rationality of gradient allocation and the coordination of the startup process. The optimization of this process not only reduces the risk of startup failure, but also improves the long-term stable operation capability of the system.

[0053] Preferably, the step of optimizing the pre-charging state in real time by using a dynamic reference ramp algorithm and an adaptive current limiting algorithm and generating power pre-charging monitoring information comprises: S21: continuously monitoring electrical reference data of each basic power supply unit in the distributed power supply in a specified form of pre-charging state; wherein the electrical reference data includes output capacitance value, load current value, and real-time feedback electrical signal; S22: Analyzing a quick startup curve of the electrical reference data of each basic power supply unit using a dynamic reference ramp algorithm and an adaptive current limiting algorithm to optimize the pre-charging specification of each basic power supply unit, and switching each basic power supply unit to a pre-charging state with the optimized pre-charging specification; S23: Arranging the electrical reference data of each basic power supply unit at each moment in time, and recording the pre-charging specification optimization operation performed by each basic power supply unit to generate power pre-charging monitoring information of each basic power supply unit.

[0054] Specifically, the electrical reference data of each basic power supply unit of the distributed power supply is continuously monitored. To perform real-time optimization of the pre-charging state, it is first necessary to continuously monitor the electrical parameters of each power supply unit. These data will provide necessary input for subsequent algorithm calculations. The electrical reference data include output capacitance value: reflecting the ability of the power supply unit to store charge during the charging process, affecting the stability and response speed of the power supply unit; load current value: indicating the actual current demand of the load, helping to determine whether the power supply unit can meet the load requirements and prevent current overload; real-time feedback electrical signal: providing real-time voltage, current and other feedback information to help accurately judge the status of the power supply unit; by continuously monitoring the electrical reference data, the change of the power supply unit status can be captured in the first time, ensuring the real-time and accuracy of the optimization process, obtaining comprehensive and accurate electrical parameter data, and being able to effectively predict the working status of the power supply unit and reduce abnormal situations during the pre-charging process.

[0055] More specifically, a fast startup curve is analyzed through a dynamic reference ramp algorithm and an adaptive current limiting algorithm. The algorithm can dynamically adjust the startup ramp according to the characteristics of the power supply unit (such as output voltage, current, and load changes) to ensure a smooth startup and prevent the power supply unit from overloading or overshooting during the startup process. By controlling the rate of change of voltage or current, sudden current fluctuations or voltage jumps are avoided to ensure system stability. The adaptive current limiting algorithm dynamically adjusts the current limit according to real-time current data to avoid damage to the power supply unit or load due to excessive current, ensuring that the current is within a safe range, and adjusting according to different power supply unit and load characteristics to provide intelligent limiting control. Through the combination of these two algorithms, the startup curve can be adjusted according to real-time electrical reference data, so that the power supply unit can start quickly without overloading. The adaptive current limiting algorithm ensures that the current will not exceed the power supply unit's tolerance, effectively avoiding damage or failure during the startup process. The dynamic reference ramp algorithm makes the startup process smoother, which not only improves the startup efficiency, but also enhances the long-term reliability of the power supply unit.

[0056] More specifically, the pre-charging specifications of each basic power supply unit are optimized so that it switches to the optimized pre-charging state. Each power supply unit needs to adjust the specifications of its charging process according to various conditions such as load, current, and heat distribution during startup. The optimization of this process helps to improve the startup efficiency of the power supply unit and avoid unnecessary loads. According to the output of the dynamic reference ramp algorithm and the adaptive current limiting algorithm, the parameters of the pre-charging process are adjusted, and the pre-charging state of the power supply unit is optimized so that it can operate under optimal conditions. By optimizing the pre-charging specifications, the startup process of the power supply unit is more stable, and faults caused by current or voltage fluctuations are reduced. The optimized pre-charging state can effectively reduce energy waste during startup and improve energy utilization efficiency. In the optimized pre-charging state, the power supply unit avoids overload or unreasonable current and voltage fluctuations, reducing the risk of equipment damage.

[0057] More specifically, the electrical reference data of each basic power supply unit is time-sequenced and the optimization operations are recorded. In order to ensure the traceability and controllability of the system, the electrical reference data of each basic power supply unit needs to be time-sequenced, which will help in subsequent analysis and optimization. Each pre-charging specification optimization operation is recorded in detail to facilitate subsequent monitoring, analysis and adjustment. By recording the timing data and optimization operations of each operation, it is ensured that the system can promptly reflect any abnormalities in the pre-charging process, thereby enhancing the maintainability of the system. The time-sequenced data can provide data support for subsequent optimization operations, ensuring that each optimization is based on the feedback of the previous operation.

[0058] More specifically, it is crucial to generate power pre-charging monitoring information in the process of real-time optimization. This monitoring information can help to timely understand the status of the power supply unit, perform fault warning and subsequent optimization. The monitoring information includes the real-time electrical parameters of the power supply unit, the pre-charging specification optimization history, the current pre-charging status, the optimized parameters and status, etc. By generating detailed monitoring information, the pre-charging status of the power supply unit can be monitored throughout the process, and potential problems can be identified in time. The recording of monitoring information provides important data support for subsequent system optimization, fault analysis and maintenance decisions.

[0059] It can be understood that through the above five steps, the pre-charging state of the distributed power supply is optimized in real time using the dynamic reference ramp algorithm and the adaptive current limiting algorithm, and the power pre-charging monitoring information is generated, which can significantly improve the startup efficiency and stability of the power supply unit. Each step in the optimization process is adjusted according to the real-time electrical data to ensure the reliability, safety and high efficiency of the system. In addition, the generation of monitoring information provides a solid data foundation for the subsequent optimization and fault warning of the system.

[0060] Preferably, the step of parsing a fast startup curve of the electrical reference data of each basic power supply unit using a dynamic reference ramp algorithm and an adaptive current limiting algorithm to optimize the pre-charging specifications of each basic power supply unit, and switching each basic power supply unit to a pre-charging state with the optimized pre-charging specifications includes: S221: performing expected simulation, risk assessment, and slope optimization of a reference voltage rising slope on the electrical reference data using a dynamic reference ramp algorithm to obtain a voltage pre-rising slope of each basic power supply unit in a currently specified pre-charging state; S222: performing a restrictive analysis of the maximum peak value of the inductor current and the load requirement of the power supply unit on the electrical reference data using an adaptive current limiting algorithm to obtain a current limiting range of each of the basic power supply units; S223: Analyzing specific quick startup execution measures for the electrical reference data of each basic power supply unit based on the voltage pre-rising slope and the current limit range, so as to obtain a quick startup curve for each basic power supply unit in a pre-charging state of a currently specified pre-charging specification; S224: Perform independent analysis and collaborative interactive analysis on the quick start curve of each of the basic power supply units to obtain optimization tendency information of the quick start curve of each of the basic power supply units, and optimize the currently specified pre-charging specifications of each of the basic power supply units based on the optimization tendency information, so that each of the basic power supply units switches to the pre-charging state of the optimized pre-charging specifications.

[0061] Specifically, the dynamic reference ramp algorithm is used to perform expected simulation, risk assessment and slope optimization of the reference voltage rising slope of the electrical reference data. The rising slope of the reference voltage determines the rate of change of the voltage of the power supply unit during the startup process. A rising slope that is too fast may cause the power supply unit to overload, and a slope that is too slow may affect the startup efficiency. The dynamic reference ramp algorithm can simulate the voltage rising process under different load conditions and predict the performance of the power supply unit based on these simulation results. By evaluating the risks under different voltage rising slopes, potential system instability factors can be identified early in the pre-charging process, such as excessive current or too fast voltage rise causing equipment overload. Optimizing the voltage rising slope can balance the startup speed and safety, thereby ensuring that the power supply unit boosts at the most appropriate rate, avoiding overload and improving efficiency. By optimizing the voltage rising slope, problems such as overload and overvoltage of the power supply unit during the startup process are avoided, and the stability of the system is enhanced. Through reasonable slope optimization, the startup time can be accelerated while ensuring that the system operates within an acceptable current and voltage range.

[0062] More specifically, the adaptive current limiting algorithm is used to perform a restrictive analysis of the electrical reference data between the maximum peak value of the inductor current and the load demand of the power supply unit. The current peak of the inductor determines the maximum current that the power supply unit can withstand during the startup process. The adaptive current limiting algorithm can be used to analyze the maximum peak value of the inductor current based on the electrical reference data to ensure that the current does not exceed the maximum carrying capacity of the power supply unit. The load demand of the power supply unit varies in different startup stages. The current limiting algorithm needs to limit the current under different load conditions to ensure that the current remains within a safe range while meeting the actual demand of the load. The adaptive current limiting algorithm ensures that the power supply unit will not be damaged due to excessive current, while avoiding the problem of insufficient load due to too small current. By adjusting the current limit in real time, the current can be dynamically adjusted according to the load demand to ensure that the power supply unit can operate efficiently and stably.

[0063] More specifically, the specific implementation measures for quick startup of the electrical reference data are analyzed based on the voltage preparation rise slope and the current limit range. The voltage rise slope and the current limit range are key factors affecting the startup process of the power supply unit. Combining the data of these two, quick startup measures can be formulated for the power supply unit, thereby optimizing the startup process. By analyzing the electrical reference data, it is determined how to increase the voltage smoothly and quickly without exceeding the current limit, ensuring quick startup and avoiding potential failures. By reasonably arranging the voltage and current adjustment strategies, the quick startup curve of the power supply unit is optimized, so that the system can start quickly while maintaining high efficiency and stability. During the quick startup process, the waste of current and voltage is avoided, and the energy utilization efficiency is improved.

[0064] More specifically, the quick start-up curve of each basic power supply unit is analyzed independently and interactively for synergy. The quick start-up curve of each basic power supply unit is analyzed separately, which can identify the performance characteristics of each power supply unit in an independent situation and ensure that each power supply unit is optimized to the best degree during the startup process. The synergy between power supply units is crucial, especially in distributed power supply systems. By analyzing the mutual influence between power supply units, the startup process can be further optimized to avoid system performance degradation due to mutual interference. Through interactive analysis, not only the startup process of a single power supply unit is optimized, but also the synergy between various parts of the system is enhanced to ensure the consistency and stability of the entire system during the startup process. The combination of independent analysis and interactive analysis ultimately optimizes the quick start-up curve of the entire system and improves the overall startup efficiency of the system.

[0065] More specifically, the pre-charging specifications are optimized based on the optimization tendency information. Through the optimization tendency information obtained in the previous steps, the pre-charging specifications can be finally optimized according to the characteristics of each basic power supply unit. The optimization tendency information reflects the performance changes of the power supply unit under different conditions, thereby providing a basis for adjusting the pre-charging specifications. According to these optimization tendency information, the pre-charging specifications can be adjusted so that each power supply unit starts in the optimal working state, thereby improving the overall efficiency and stability of the system. Through the optimized pre-charging specifications, each power supply unit can be started in the optimal working state, thereby significantly improving the overall performance of the system. By adjusting the pre-charging specifications, failure or damage of the power supply unit caused by parameter mismatch during the startup process is avoided.

[0066] More specifically, switch to the pre-charging state of the optimized pre-charging specification. After the previous steps are completed, the power supply unit can switch to the optimized pre-charging specification to achieve the optimal startup process and working state. By smoothly switching to the optimized pre-charging specification, it can be ensured that the power supply unit transitions to a stable working state, while ensuring that the power supply unit is not affected by excessive load. After switching to the optimized pre-charging specification, the power supply unit can operate in a more stable and efficient state, reducing potential failures. The optimized pre-charging specification provides the power supply unit with the best working environment, thereby improving the overall operating efficiency of the system.

[0067] It is understandable that through this series of steps, from the dynamic reference ramp algorithm to the adaptive current limiting algorithm, and then to the analysis and optimization of specific implementation measures, it can ultimately ensure that the distributed power supply system is both efficient and stable during startup. The comprehensive application of these steps improves the rapid startup efficiency of the power supply unit, reduces the possible failure risk during the startup process, and provides the power supply unit with the optimal working state, thereby improving the performance and reliability of the entire system.

[0068] Preferably, the steps of performing a time-frequency domain joint analysis on the power pre-charging monitoring information and substituting the analysis results into a lightweight CNN model deployed at the edge to generate power load behavior characteristics include: S31: performing time domain analysis of numerical statistical feature extraction and event detection and frequency domain analysis of fast Fourier transform and wavelet transform on the electrical reference data arranged in time sequence in the power pre-charging monitoring information, and vectorizing the time domain analysis results and the frequency domain analysis results to generate a time domain feature matrix and a frequency domain feature matrix; S32: constructing an adjacency matrix of the time domain feature matrix and the frequency domain feature matrix, and performing cluster analysis on the adjacency matrix, so as to extract and combine key information of the time domain feature matrix and the frequency domain feature matrix according to the result of the cluster analysis, so as to obtain a time-frequency domain joint spectrum; S33: Feature encoding is performed on the pre-charging specification optimization operations recorded in the power pre-charging monitoring information to embed them into corresponding positions in the time-frequency domain joint map, and the time-frequency domain joint map is substituted into the pre-trained lightweight CNN model deployed on the edge device, so as to parse the time-frequency domain joint map through the lightweight CNN model to generate power load behavior characteristics.

[0069] Specifically, the electrical reference data in the power pre-charging monitoring information is analyzed in the time domain and frequency domain. The key to time domain analysis is to extract statistical features and detect events in the electrical reference data during the power pre-charging process, which can help identify abnormal behaviors or features in the pre-charging process, such as excessive current, excessive voltage, etc. Numerical statistical feature extraction, such as mean, standard deviation, maximum value, minimum value, etc., is used to describe the basic behavior of electrical data. Event detection of electrical data helps to identify key events (such as sudden changes in voltage and current) that occur during the pre-charging process; Fast Fourier Transform (FFT) and wavelet transform are used to Frequency domain analysis is performed on electrical data to identify signal characteristics within different frequency ranges. Fast Fourier Transform (FFT) can reveal the periodic components and frequency characteristics in electrical signals, helping to identify the performance of power loads at different frequencies. Wavelet transform is suitable for analyzing non-stationary signals and can reveal local characteristics in time and frequency. It has good analytical capabilities for instantaneous changes and sudden events in electrical signals. Through joint analysis of the time domain and frequency domain, the key characteristics of power load behavior can be more comprehensively captured. The generation of time domain feature matrices and frequency domain feature matrices provides more detailed and comprehensive data support for subsequent analysis.

[0070] More specifically, the time domain analysis and frequency domain analysis results are vectorized to generate time domain and frequency domain feature matrices. The numerical statistical features extracted from the time domain analysis results are used to form a feature matrix that can represent the time domain behavior of the power supply pre-charging process. The frequency domain analysis results, especially the results of FFT and wavelet transform, are used to generate frequency domain feature matrices. These matrices can reflect the frequency components of the power supply load behavior. By representing the time domain and frequency domain features in a matrix, subsequent calculations and model inputs can be performed more conveniently, and data processing efficiency can be improved. Representing the time domain and frequency domain data as matrices respectively can effectively retain their respective feature information and prepare for subsequent joint analysis.

[0071] More specifically, an adjacency matrix of the time domain feature matrix and the frequency domain feature matrix is ​​constructed and cluster analysis is performed. The adjacency matrix is ​​used to represent the correlation between the time domain feature matrix and the frequency domain feature matrix. By constructing the adjacency matrix, the interaction and mutual influence between the time domain and frequency domain features can be captured. The adjacency matrix is ​​analyzed using a clustering algorithm to identify regions or patterns with similar features in the data. Cluster analysis can help screen out key feature combinations and help further extract key information. The adjacency matrix and cluster analysis can reveal the potential correlation between time domain and frequency domain features, improve the accuracy of feature extraction, and cluster analysis can help screen out representative information from massive features and optimize the input data of subsequent models.

[0072] More specifically, key information of the time domain feature matrix and the frequency domain feature matrix is ​​extracted and combined to obtain a joint time-frequency domain map. Based on the results of cluster analysis, key information in the time domain and frequency domain matrices is extracted, and features are combined to form a comprehensive joint time-frequency domain map. By merging the time domain and frequency domain information into one map, the time domain and frequency domain characteristics of the power pre-charging process can be reflected at the same time, which is convenient for subsequent analysis and model training. The joint time-frequency domain map can more comprehensively characterize the power load behavior characteristics and provide richer feature information for the model input. The fusion of time domain and frequency domain information helps to improve the accuracy of the subsequent lightweight CNN model in analyzing and predicting the power load behavior.

[0073] More specifically, the pre-charging specification optimization operation is feature-encoded and embedded in the joint time-frequency domain graph. The power supply pre-charging specification optimization operation is encoded and converted into a form that can be understood by the model and embedded in the joint time-frequency domain graph. This operation enables the model to process more diverse information, such as operation history or optimization parameters. By embedding the characteristics of the optimization operation into the graph, it is ensured that the time-frequency domain graph not only contains the basic behavioral characteristics of the power supply load, but also includes operation optimization information, which helps the model to fully understand and analyze the power supply behavior. By encoding and embedding the pre-charging specification optimization operation, the model can comprehensively analyze the power supply behavior and its optimization measures, thereby improving the prediction accuracy of the load behavior characteristics. Embedding the optimization information of the pre-charging operation into the graph helps to improve the model's adaptability and intelligent analysis capabilities to different power supply load behaviors.

[0074] More specifically, the time-frequency domain joint map is substituted into the lightweight CNN model for analysis to generate power load behavior characteristics. The pre-trained lightweight convolutional neural network (CNN) model is used to process the time-frequency domain joint map to automatically extract power load behavior characteristics. By parsing the time-frequency domain joint map through the CNN model, valuable power load behavior characteristics can be automatically learned from the map, and data support can be provided for subsequent applications. The CNN model can automatically extract the behavior characteristics of the power load from the time-frequency domain joint map without manual intervention, thereby improving processing efficiency and accuracy. By parsing the time-frequency domain joint map, the CNN model can generate accurate characteristics of the power load behavior, providing accurate data for subsequent power management and optimization.

[0075] It can be understood that this process can comprehensively and accurately analyze the power load behavior characteristics by combining the joint analysis of time and frequency domains with the analytical capabilities of the lightweight CNN model. This method not only improves the accuracy of feature extraction, but also effectively combines time domain, frequency domain and operation optimization information, thereby improving the intelligence level of power behavior analysis and optimizing the management and operation efficiency of the power system.

[0076] Preferably, the steps of reconstructing the designated form of the pre-charge state of the distributed power supply according to the power load behavior characteristics and synchronously adjusting the fast startup strategy include: S41: analyzing the stability of each of the multiple groups of parallel-connected basic power supply units in the distributed power supply according to the power load behavior characteristics, so as to generate a stability characteristic of each of the basic power supply units relative to a currently specified pre-charging specification; S42: configuring a pre-charging specification adjustment factor set for each of the basic power supply units according to the stability characteristics of the status of each of the basic power supply units; wherein the adjustment factor set includes a plurality of adjustment factors, and the adjustment factors are used to control the pre-charging specification of the basic power supply unit for adjustment; S43: Analyzing the collaborative execution effect of the adjustment factor sets of each basic power supply unit to obtain the collaborative execution value between the adjustment factors in each adjustment factor set, and generating a pre-charging reconfiguration plan for the distributed power supply based on the collaborative execution value of each adjustment factor set; S44: According to the pre-charging reconstruction plan, the pre-charging specifications of each basic power supply unit in the distributed power supply are adjusted accordingly, so that each basic power supply unit switches to the pre-charging state of the adjusted pre-charging specifications, and at the same time, the quick start strategy is synchronously adjusted according to the adjusted pre-charging specifications of each basic power supply unit.

[0077] Specifically, the status stability of each basic power supply unit is analyzed to generate status stability characteristics. According to the power load behavior characteristics, the stability of each parallel basic power supply unit in the distributed power supply is analyzed. The stability of the power supply unit is closely related to its working state (such as voltage and current fluctuations). According to the load behavior of the power supply unit, the stability characteristics of each basic power supply unit are generated to compare with the current pre-charging specifications. These characteristics reflect the stability of the power supply unit under a specific load, ensuring that potential instability or failure is avoided during the pre-charging process. Through the generation of status stability characteristics, the operating status of each basic power supply unit can be accurately evaluated, providing data support for subsequent adjustments. By evaluating the stability of the power supply unit, potential power instability risks can be discovered in advance to ensure the reliability and safety of the system.

[0078] More specifically, a set of adjustment factors for pre-charging specifications is configured for each basic power supply unit. Based on the stability characteristics of the power supply unit, a corresponding pre-charging specification adjustment factor is configured for each basic power supply unit. The configuration of the adjustment factor set ensures that the power supply unit can be adjusted according to the stability characteristics under different load conditions to achieve the optimal pre-charging state. The adjustment factors include adjustment items such as current, voltage, and charging rate, which are used to control changes in the pre-charging specifications of the power supply unit. Through the adjustment factor set, the pre-charging behavior of each power supply unit can be controlled in detail to ensure the best charging effect under different load conditions. The adjustment factor is customized according to the specific conditions of the power supply unit, thereby improving the adaptability and flexibility of the entire system.

[0079] More specifically, the collaborative execution effect of the adjustment factor set is analyzed to generate a pre-charging reconstruction plan, the collaborative execution value between each adjustment factor is analyzed, and their synergy in practical applications is evaluated. Because the adjustment of multiple power units may produce interactive effects, only by analyzing their collaborative execution value can we ensure that the effect of each adjustment factor can be maximized. Based on the collaborative execution value between each factor, an overall pre-charging reconstruction plan is formulated to ensure that all power units can work stably and efficiently after adjustment. Through collaborative analysis, the best adjustment strategy can be identified to avoid system instability or reduced efficiency due to uncoordinated adjustment of individual power units. By coordinating the execution effects of adjustment factors, the operating efficiency and stability of the entire distributed power system can be improved.

[0080] More specifically, the pre-charging specifications of the basic power supply unit are adjusted according to the pre-charging reconstruction plan. According to the pre-charging reconstruction plan, the pre-charging specifications of each basic power supply unit in the distributed power supply are adjusted to ensure that each power supply unit can be pre-charged in the most optimized manner to achieve the best working state. According to the adjusted pre-charging specifications, the power supply unit will switch to a new working mode to ensure its optimal performance in terms of stability and load characteristics. Through precise adjustment of the pre-charging specifications, the performance and stability of the power supply unit can be maximized, and failures or performance degradation caused by improper charging can be reduced. The pre-charging state of each power supply unit is personalized, which improves the adaptability and flexibility of the distributed power supply system.

[0081] More specifically, the quick start strategy is adjusted synchronously. According to the adjusted pre-charging specifications, the quick start strategy also needs to be adjusted synchronously. The quick start strategy refers to how to quickly and stably switch the power supply from the startup state to the normal working state during the startup of the power supply system. By synchronously adjusting the startup strategy, a smooth transition can be ensured during the power pre-charging and startup process. The startup parameters include startup time, startup force, startup current and other parameters are adjusted to ensure that there will be no excessive fluctuations during the startup process. By synchronously adjusting the quick start strategy, overload or instability of the power supply unit during the startup process can be avoided, ensuring that the power supply system starts quickly and smoothly. By precisely controlling the parameters during the startup process, the response speed and reliability of the power supply system are improved, especially in the face of load changes or abnormal conditions.

[0082] It can be understood that through this series of steps, the power load behavior characteristics are combined with the pre-charging state of the distributed power supply and the adjustment of the fast startup strategy, which can achieve dynamic optimization of the power supply system. This process effectively improves the stability and efficiency of the power supply unit and the startup response speed of the system, ensuring the reliable operation of the system in complex and dynamic environments.

[0083] Preferably, when a power supply startup instruction is monitored, the step of quickly starting the distributed power supply according to the fast startup strategy includes: S51: when a power startup instruction is monitored, performing a priority startup operation based on a pre-calculated fast startup curve on each basic power supply unit as a priority gradient in the distributed power supply according to the fast startup strategy; S52: performing a startup operation based on a pre-calculated fast startup curve on each basic power supply unit of the distributed power supply as a normal gradient according to the fast startup strategy.

[0084] Specifically, the power startup instruction is monitored and a quick startup is executed. When the system detects a power startup instruction, the quick startup process is started immediately. This step is the trigger condition for the startup process, ensuring that the system can respond quickly when needed and can respond to the startup instruction in real time, thereby improving the flexibility and quick response capability of the system.

[0085] More specifically, priority startup operations are performed on power supply units with priority gradients. For power supply units with priority in the quick startup process (such as core power supply units that need to be stabilized as soon as possible), startup operations are performed according to a pre-calculated quick startup curve. The quick startup curve is derived by analyzing the power load characteristics and the behavior patterns during the startup process to ensure that each power supply unit can be started in the most optimized manner. The startup path calculated for each basic power supply unit takes into account multiple factors such as current, voltage, and time during startup. It aims to achieve a fast and stable startup state by minimizing possible load fluctuations during the startup process. By performing startup operations according to the preset quick startup curve, it can be ensured that the power supply units with priority gradients can be quickly and stably put into working state, avoiding failures caused by excessive load or instability.

[0086] More specifically, normal startup operations are performed on power supply units with normal gradients. For power supply units with lower priorities, the system will perform startup operations according to the same quick startup curve. These power supply units have lower startup priorities, but still need to be started according to the optimized startup curve to ensure that they are put into operation normally, ensure that the power supply units with normal gradients are started at the appropriate time, avoid load fluctuations during the startup process, and at the same time will not affect the startup efficiency of the priority gradient power supply units. By synchronously adjusting the priority and startup operations, it is ensured that all power supply units can operate in a balanced manner and complete stable startup in a short time.

[0087] It can be understood that through this fast start-up strategy, the distributed power supply system can perform optimized startup operations for priority gradient and ordinary gradient power supply units according to the priority of the power supply unit. This strategy is based on a pre-calculated fast start-up curve to ensure that the power supply system can start quickly and stably after receiving the startup instruction, while maximizing the startup efficiency and system stability.

[0088] Reference Figure 2 As shown, in a second aspect, the present invention provides a rapid startup device for an onboard power module, which is used to implement a rapid startup method for an onboard power module according to any one of the first aspects, comprising: A pre-start module is used to pre-charge the distributed power supply pre-installed on the circuit board and deploy a corresponding fast start strategy for the distributed power supply in a specified pre-charge state; A real-time optimization module, configured to optimize the pre-charging state in real time by using a dynamic reference ramp algorithm and an adaptive current limiting algorithm, and to generate power pre-charging monitoring information; A feature monitoring module is used to perform a joint time-frequency domain analysis on the power pre-charging monitoring information and substitute the analysis results into a lightweight CNN model deployed at the edge to generate power load behavior characteristics; a layout reconstruction module, configured to reconstruct a specified form of a pre-charge state of the distributed power supply according to the power load behavior characteristics, and synchronously adjust the fast startup strategy; The fast startup module is used to quickly start the distributed power supply according to the fast startup strategy when a power startup instruction is monitored.

[0089] In this embodiment, for the specific implementation of each module in the above system embodiment, please refer to the above method embodiment, which will not be repeated here.

[0090] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A quick startup method for a board-mounted power module, characterized in that: include: Perform pre-charging processing on the distributed power supply pre-installed on the circuit board, and deploy corresponding fast startup strategies for the distributed power supply in the specified pre-charging state; The pre-charging state is optimized in real time by a dynamic reference ramp algorithm and an adaptive current limiting algorithm, and power supply pre-charging monitoring information is generated; Performing a joint time-frequency domain analysis on the power pre-charge monitoring information, and substituting the analysis results into a lightweight CNN model deployed at the edge to generate power load behavior characteristics; Reconstructing a specified form of the pre-charge state of the distributed power supply according to the power load behavior characteristics, and synchronously adjusting the fast startup strategy; When a power startup instruction is monitored, the distributed power supply is quickly started according to the fast startup strategy.

2. The method for quickly starting a board-mounted power module according to claim 1, wherein: The steps of pre-charging the distributed power supply pre-set on the circuit board and deploying a corresponding fast startup strategy for the distributed power supply in a specified pre-charge state include: Acquiring power supply structure information of a distributed power supply pre-set on a circuit board, so as to allocate unit connection marks for describing electrical connection relationships to each of a plurality of parallel-connected basic power supply units in the distributed power supply; Based on the power supply configuration information of the distributed power supply, a quick startup gradient analysis is performed on each of the basic power supply units to assign a startup gradient mark to each of the basic power supply units; Specifying a pre-charging specification for each basic power supply unit according to a startup gradient tag assigned to each basic power supply unit, so that each basic power supply unit is pre-charged in accordance with the specified pre-charging specification through a pre-charging circuit, so that each basic power supply unit is in a pre-charging state in accordance with the specified pre-charging specification, thereby placing the distributed power supply in a specified form of pre-charging state; Based on the startup gradient mark and unit connection mark assigned to each of the basic power supply units, as well as the pre-charging specifications of each of the basic power supply units, a priority control architecture analysis for rapid power startup is performed on each of the basic power supply units to generate a rapid startup strategy for the distributed power supply in a specified form of pre-charging state.

3. The method for quickly starting a board-mounted power module according to claim 2, wherein: The step of performing a quick startup gradient analysis on each of the basic power supply units based on the power supply configuration information of the distributed power supply to assign a startup gradient mark to each of the basic power supply units includes: Based on the power supply structure information, digital simulation is performed on the distributed power supply to determine electrical parameter correlation, electrical connection topology, and thermal distribution. The digital simulation results are used as multiple constraints to evaluate the adaptability of each basic power supply unit in the distributed power supply as a primary pre-charging target, thereby obtaining adaptability characteristics of each basic power supply unit as a primary pre-charging target. According to the adaptability characteristics of each of the basic power supply units as a main pre-charging object, the distributed power sources are preliminarily allocated with priority gradients and common gradients as an original gradient division form; The original gradient division form is evaluated for value through a pre-built distributed power supply rapid startup knowledge base, so as to iteratively optimize the gradient allocation of the original gradient division form, and to allocate startup gradient marks to each of the basic power supply units based on the optimized allocation form.

4. The method for quickly starting a board-mounted power module according to claim 2, wherein: The steps of optimizing the pre-charging state in real time by using a dynamic reference ramp algorithm and an adaptive current limiting algorithm and generating power pre-charging monitoring information include: Continuously monitoring electrical reference data of each basic power supply unit in the distributed power supply in a specified form of pre-charge state; wherein the electrical reference data includes output capacitance value, load current value, and real-time feedback electrical signal; Performing a quick startup curve analysis on the electrical reference data of each basic power supply unit using a dynamic reference ramp algorithm and an adaptive current limiting algorithm to optimize the pre-charging specifications of each basic power supply unit, so that each basic power supply unit switches to a pre-charging state with the optimized pre-charging specifications; The electrical reference data of each basic power supply unit at each moment is time-series-arranged, and the pre-charging specification optimization operation performed by each basic power supply unit is recorded to generate power pre-charging monitoring information of each basic power supply unit.

5. The method for quickly starting a board-mounted power module according to claim 4, wherein: The steps of analyzing the electrical reference data of each basic power supply unit for a quick startup curve using a dynamic reference ramp algorithm and an adaptive current limiting algorithm to optimize the pre-charging specifications of each basic power supply unit, and switching each basic power supply unit to a pre-charging state with the optimized pre-charging specifications include: Performing expected simulation, risk assessment, and slope optimization of a reference voltage rising slope on the electrical reference data using a dynamic reference ramp algorithm to obtain a voltage pre-rising slope of each basic power supply unit in a currently specified pre-charging state; Performing a restrictive analysis of the maximum peak value of the inductor current and the load requirement of the power supply unit on the electrical reference data using an adaptive current limiting algorithm to obtain a current limiting range of each of the basic power supply units; Analyzing specific quick startup execution measures for electrical reference data of each basic power supply unit based on the voltage pre-rising slope and the current limit range to obtain a quick startup curve for each basic power supply unit in a pre-charging state of a currently specified pre-charging specification; The quick start-up curves of each of the basic power supply units are independently analyzed and interactively analyzed for synergy to obtain optimization tendency information of the quick start-up curves of each of the basic power supply units, and the currently specified pre-charging specifications of each of the basic power supply units are optimized based on the optimization tendency information so that each of the basic power supply units switches to the pre-charging state of the optimized pre-charging specifications.

6. The method for quickly starting a board-mounted power module according to claim 4, wherein: The steps of performing a joint time-frequency domain analysis on the power pre-charging monitoring information and substituting the analysis results into a lightweight CNN model deployed at the edge to generate power load behavior characteristics include: Performing time domain analysis of numerical statistical feature extraction and event detection and frequency domain analysis of fast Fourier transform and wavelet transform on the electrical reference data arranged in time sequence in the power pre-charging monitoring information, and vectorizing the time domain analysis results and the frequency domain analysis results to generate a time domain feature matrix and a frequency domain feature matrix; Constructing an adjacency matrix of the time domain feature matrix and the frequency domain feature matrix, and performing cluster analysis on the adjacency matrix, so as to extract and combine key information of the time domain feature matrix and the frequency domain feature matrix according to the results of the cluster analysis, so as to obtain a time-frequency domain joint spectrum; Feature encoding is performed on the pre-charging specification optimization operations recorded in the power pre-charging monitoring information to embed them into corresponding positions in the time-frequency domain joint map, and the time-frequency domain joint map is substituted into a pre-trained lightweight CNN model deployed on the edge device, so as to parse the time-frequency domain joint map through the lightweight CNN model to generate power load behavior characteristics.

7. The method for quickly starting a board-mounted power module according to claim 2, wherein: The steps of reconstructing the specified form of the pre-charge state of the distributed power supply according to the power load behavior characteristics and synchronously adjusting the fast startup strategy include: Analyzing the stability of each of the multiple parallel-connected basic power supply units in the distributed power supply according to the power load behavior characteristics to generate a stability characteristic of each of the basic power supply units relative to a currently specified pre-charging specification; configuring a pre-charge specification adjustment factor set for each basic power supply unit according to the condition stability characteristics of each basic power supply unit; wherein the adjustment factor set includes a plurality of adjustment factors, and the adjustment factors are used to control the pre-charge specification of the basic power supply unit for adjustment; Analyzing the collaborative execution effect of the adjustment factor sets of each of the basic power supply units to obtain the collaborative execution value between the adjustment factors in each of the adjustment factor sets, and generating a pre-charging reconfiguration plan for the distributed power supply based on the collaborative execution value of each of the adjustment factor sets; According to the pre-charging reconstruction plan, the pre-charging specifications of each basic power supply unit in the distributed power supply are adjusted accordingly so that each basic power supply unit switches to the pre-charging state of the adjusted pre-charging specifications. At the same time, the quick start strategy is synchronously adjusted according to the adjusted pre-charging specifications of each basic power supply unit.

8. The method for quickly starting a board-mounted power supply module according to claim 3, wherein: When a power supply startup instruction is monitored, the step of quickly starting the distributed power supply according to the quick startup strategy includes: When a power startup instruction is monitored, a priority startup operation is performed on each basic power supply unit as a priority gradient in the distributed power supply according to the fast startup strategy based on a pre-calculated fast startup curve; According to the fast startup strategy, a startup operation based on a pre-calculated fast startup curve is performed on each basic power supply unit of the distributed power supply as a normal gradient.

9. A quick start device for a board-mounted power module, characterized in that: A method for quickly starting a board-mounted power module according to any one of claims 1 to 8, comprising: A pre-start module is used to pre-charge the distributed power supply pre-installed on the circuit board and deploy a corresponding fast start strategy for the distributed power supply in a specified pre-charge state; A real-time optimization module, configured to optimize the pre-charging state in real time by using a dynamic reference ramp algorithm and an adaptive current limiting algorithm, and to generate power pre-charging monitoring information; A feature monitoring module is used to perform a joint time-frequency domain analysis on the power pre-charging monitoring information and substitute the analysis results into a lightweight CNN model deployed at the edge to generate power load behavior characteristics; a layout reconstruction module, configured to reconstruct a specified form of a pre-charge state of the distributed power supply according to the power load behavior characteristics, and synchronously adjust the fast startup strategy; The fast startup module is used to quickly start the distributed power supply according to the fast startup strategy when a power startup instruction is monitored.

Images