Health adaptive control method of guideway power supply under multi-load access condition

Abstract

The present application relates to the field of adaptive control and state monitoring, in particular to a health adaptive control method of a guide rail type power supply under the condition of multiple load access; the method comprises: collecting load access transient electrical parameters in real time and constructing a structured behavior sequence reflecting impact intensity, duration and recovery characteristics, and then mapping the behavior sequence to equivalent stress input and using a recursive evolution method to infer the internal electrical stress and thermal stress state of the power supply, on this basis, short-term analysis and long-term trend smoothing processing are carried out on the dual-component health state, and a multi-dimensional risk level is divided, and finally the health constraint judgment of load access, power flexible adjustment and dynamic priority scheduling of multiple load access are realized according to the health state and risk level. The present application forms a closed-loop management mechanism from behavior perception to internal state inference to health driven control, which improves the reliability and life friendliness of the power supply system under the condition of multiple load random access.

Description

Technical Field

[0001] This invention relates to the field of adaptive control and condition monitoring technology, specifically to a health adaptive control method for rail-mounted power supplies under multi-load conditions. Background Technology

[0002] With the rapid development of industrial automation, the Internet of Things and distributed device systems, DIN rail power supplies are widely used in various industrial control and embedded scenarios due to their compact, modular and easy-to-deploy characteristics. In actual operation, these power supplies often face complex working conditions with multiple loads randomly connected and dynamically changing, which puts forward higher requirements for the instantaneous response capability, long-term operational stability and life management of the power supply system.

[0003] Chinese invention patent application CN121324970A discloses a method and system for health assessment of lead-acid batteries based on simulated operating conditions. The method includes: collecting simulated operating data of lead-acid batteries and cleaning the data to obtain high-quality battery data; filtering recovery effect data and removing it from the high-quality battery data to obtain disturbance-removed battery data; performing polarization voltage compensation to generate voltage-compensated battery data; performing abnormal impact identification and filtering abnormal data from the voltage-compensated battery data based on the identification results to output reasonable battery data; identifying long-cycle data and performing risk assessment to obtain corrected battery data and covering the corresponding data in the reasonable battery data to obtain differentiated battery data; and assessing the health status of the differentiated battery data to output lead-acid battery health indicators.

[0004] Currently, fault prediction and health management technologies are playing an increasingly important role in the operation and maintenance of critical equipment. Their core lies in the transformation from routine maintenance to proactive health assurance through continuous monitoring, evaluation, and prediction of equipment status. Against this backdrop, how to enable DIN rail power supplies not only to reliably cope with transient impacts from multiple load connections, but also to perform real-time quantitative assessment and adaptive adjustment of their own health status, thereby improving the overall reliability of the system and achieving intelligent lifespan management, has become an important evolutionary direction and development trend in the field of power management technology. Summary of the Invention

[0005] The purpose of this invention is to address the problems existing in the background technology by proposing a health adaptive control method for rail-mounted power supplies under multi-load conditions.

[0006] The technical solution of this invention: a health adaptive control method for a rail-mounted power supply under multi-load conditions, comprising the following specific implementation steps: S1. Real-time acquisition of the output current and output voltage of the rail-mounted power supply under multi-load random access conditions, identification of load access transient events, extraction and construction of a structured load access behavior vector reflecting the access impact intensity, disturbance duration and recovery characteristics based on the transient sequence of output current, and formation of a multi-load access behavior sequence in chronological order. S2. The load access behavior vector in the multi-load access behavior sequence is mapped to the equivalent stress input through weighted combination. The internal comprehensive stress state of the rail-mounted power supply is dynamically updated and inferred using a recursive evolution method. The internal comprehensive stress state is decomposed into electrical stress components and thermal stress components. After normalization, an internal stress state observation sequence is formed. S3. Based on the internal stress state observation sequence, the initial health status index with two components is calculated. The initial health status index is subjected to short-term change analysis and long-term trend smoothing to obtain the long-term health status index. The multi-dimensional risk level is then divided in combination with the health trend index. S4. Using long-term health status indicators, health trend indicators, and multi-dimensional risk levels as constraints, the system performs health constraint judgment and dynamic priority scoring on the loads applying for access. For loads that are determined to be allowed to access, the system performs flexible power adjustment based on the health margin. After the load access is completed, the system performs closed-loop updates on the internal stress state and health state to achieve dynamic optimization scheduling of multiple load access.

[0007] Preferably, in step S1, the specific process of constructing the load access behavior vector includes: The start time of the load connection transient event is determined by using the rate of change of the output current exceeding a threshold determined based on the statistical characteristics of steady-state operation. The event ends when the output current changes back to the steady-state range. Based on the transient sequence of the output current, the rate of change of the current from the start time to the end time is integrated to calculate the index characterizing the intensity of the access impact. An index characterizing the duration of the disturbance is calculated based on the start and end times; Based on the time required for the output current to reach a steady-state value after the load connection event ends, an index characterizing the recovery characteristics is calculated. The access impact intensity index, disturbance duration index, and recovery characteristic index calculated for the same load access event are combined and encapsulated to form the load access behavior vector for that event.

[0008] Preferably, in step S2, the specific process of the recursive evolution method is as follows: The three dimensions of the load access behavior vector—access impact intensity, disturbance duration, and recovery characteristics—are weighted and combined according to the weighting coefficients calibrated based on the electrical and thermal characteristics of the power supply devices, and mapped to the equivalent stress input corresponding to the load access for the kth time. The internal comprehensive stress state after the kth load connection is obtained by adding the equivalent stress input of the kth load connection and the internal comprehensive stress state after the (k-1)th load connection by the stress attenuation coefficient. The internal stress state can be further decomposed into electrical stress components and thermal stress components: The electrical stress component is obtained by adding the current equivalent stress input's impact intensity portion to the previous electrical stress component multiplied by the electrical stress attenuation coefficient; The thermal stress component is obtained by adding another part of the current equivalent stress input to the value of the previous thermal stress component multiplied by the thermal stress attenuation coefficient.

[0009] Preferably, in step S3, the specific calculation process for the initial health status index of the two components is as follows: The normalized electrical stress observation values ​​output in step S2 are converted into electrical stress health indicators through the first preset mapping relationship. The normalized thermal stress observation values ​​are converted into thermal stress health indicators through a second preset mapping relationship; The electrical stress health index and the thermal stress health index together constitute a two-component initial health state index that reflects the power supply's ability to withstand transient impacts and the level of cumulative thermal load.

[0010] Preferably, in step S3, the long-term trend smoothing process specifically employs an exponentially weighted moving average method, specifically as follows: The initial health status index sequence arranged in chronological order is processed, where the long-term health status index value at the current moment is obtained by multiplying the initial health status index value at the current moment by a smoothing coefficient, and adding the difference between multiplying the long-term health status index value at the previous moment by a factor minus the smoothing coefficient. The smoothing coefficient ranges from zero to one and is used to adjust the degree of impact of short-term fluctuations on long-term health trends.

[0011] Preferably, in step S3, the specific process of classifying multidimensional risk levels includes: The electrical stress health component and thermal stress health component in the long-term health status index are compared with the electrical stress safety health threshold, thermal stress safety health threshold, electrical stress warning threshold, thermal stress warning threshold, electrical stress danger threshold and thermal stress danger threshold calibrated based on the rated withstand capacity and safety margin of the device. Calculate the health trend index, which represents the amount of change in long-term health status within a set historical window. Based on the threshold range of the electrical stress health component and the thermal stress health component, as well as whether the health trend indicator is declining, the risk level is comprehensively determined as green safety level, yellow warning level, orange risk level, and red danger level.

[0012] Preferably, in step S4, the specific process for determining health constraints is as follows: Obtain the transient impact characteristic vector of the load to be connected. This vector includes the estimated access impact intensity, disturbance duration, and recovery characteristics when the load is connected. Read the electrical stress health component and thermal stress health component from the long-term health status indicators at the current moment; Determine whether the electrical stress health component is higher than the electrical stress safety lower limit threshold, whether the thermal stress health component is higher than the thermal stress safety lower limit threshold, and whether the load's preset access priority is higher than the set minimum allowable threshold. The health constraint decision is to allow access only if all judgment conditions are met; otherwise, the decision is to delay access and enter the power regulation process.

[0013] Preferably, in step S4, the specific process of power flexible adjustment is as follows: When the load health constraint determination result is that it needs to enter the power regulation process, the electrical stress health margin is determined based on the difference between the current electrical stress health component and the electrical stress safety lower limit threshold, and the thermal stress health margin is determined based on the difference between the current thermal stress health component and the thermal stress safety lower limit threshold. By combining the mapping coefficients of electrical stress and thermal stress in the transient impact characteristic vector of the load, the adaptive access power is dynamically calculated. The adaptive access power is less than the rated operating power of the load, and its value is positively correlated with the electrical stress health margin and the thermal stress health margin.

[0014] Preferably, in step S4, the specific process of closed-loop update is as follows: After the load is actually connected according to the rated power and adaptive connection power, the actual contribution of this connection to electrical stress and thermal stress is calculated based on the ratio of the actual connection power to the rated power. The actual contribution is superimposed onto the electrical stress component and thermal stress component before the update, and then multiplied by the corresponding electrical stress time decay coefficient and thermal stress time decay coefficient respectively to obtain the updated electrical stress component and thermal stress component. Based on the updated electrical stress and thermal stress components, normalization and health status index calculations are performed again to complete the closed-loop update of long-term health status indicators.

[0015] Preferably, in step S4, the specific process of generating the dynamic priority score is as follows: When several loads simultaneously request access, a dynamic priority score is calculated for each load. The calculation of this score takes into account the correlation between the current power supply’s long-term electrical stress health level and the load’s estimated electrical shock characteristics, the correlation between the long-term thermal stress health level and the load’s estimated thermal shock characteristics, and the magnitude of the load’s own power demand. Among them, the estimated electrical and thermal shock characteristics of the load have a negative impact on its score; During the calculation, weight coefficients determined by the system operation strategy are assigned to the electrical stress health correlation item, thermal stress health correlation item, and power demand item respectively; The loads applying for access are sorted according to the calculated dynamic priority score. Loads with higher scores are given priority to access permission, while loads with lower scores are delayed in access or enter the power adjustment process based on the real-time updated health status, thus forming a closed-loop dynamic access scheduling.

[0016] Compared with the prior art, the above-mentioned technical solution of the present invention has the following beneficial technical effects: This invention designs a health adaptive control method for rail-mounted power supplies under multi-load conditions. By behaviorally representing and serializing the transient process of load connection, it achieves correlation modeling between external disturbances and internal states of the power supply, transforming originally discrete and random load connection events into continuous and analyzable system inputs. Based on this, through equivalent stress mapping and recursive evolution mechanisms, it realizes real-time inference and quantitative monitoring of the internal electrical and thermal stress states of the power supply, overcoming the limitation of relying solely on external parameters such as output voltage and current, which cannot reflect the true internal stress state. Based on the dual-component health state index and its synergistic analysis of short-term changes and long-term trends, the system can effectively distinguish between recoverable fluctuations caused by transient impacts and performance degradation caused by cumulative degradation, thus achieving accurate monitoring of the power supply's health state. Dynamic and continuous assessment provides a reliable basis for fault prediction and health management. Ultimately, through an adaptive adjustment mechanism centered on health status and risk level, the system can dynamically optimize load access decisions and power allocation in scenarios with multiple concurrent loads. This not only suppresses sudden impacts from high-risk load access but also enables soft load access through flexible power adjustment. Thus, while ensuring the normal operation of the system, it significantly reduces the comprehensive electrothermal stress borne by the power supply, slows down the aging process of components, and improves the long-term reliability, stability, and lifespan friendliness of the system. Overall, this invention realizes a paradigm shift in rail-mounted power management from passive response and fixed threshold protection to active sensing, state inference, and health-driven adaptive control, improving the intelligent operation and maintenance level of power systems under complex multi-load conditions. Attached Figure Description

[0017] Figure 1This is a flowchart of a health adaptive control method for a rail-mounted power supply under multi-load conditions proposed in this invention. Detailed Implementation

[0018] Example 1, as Figure 1 As shown, the present invention proposes a health adaptive control method for a rail-mounted power supply under multi-load conditions, which includes the following specific implementation steps: S1. By finely sensing the transient changes of the output electrical parameters of the rail-mounted power supply under the condition of random multi-load connection, the load connection process is effectively separated from the continuous operation background, and the connection process is quantitatively described in a behavioral rather than steady-state manner. Then, a structured behavioral sequence that can reflect the connection impact intensity, disturbance persistence characteristics and recovery capability is constructed. The specific implementation process is as follows: S11. By continuously sampling the output current and output voltage of the rail-mounted power supply for a period longer than the regulation cycle, using the current change rate as a trigger to identify transient load connection events, and opening the transient acquisition window when the event occurs, the acquired data is aligned to a unified time reference. This allows for the reliable capture of comparable transient electrical parameter data of the load connection under complex multi-load operating conditions. Specifically: At the output end of the rail-mounted power supply, the output current and output voltage are continuously sampled. The sampling frequency is set higher than the normal regulation frequency of the power supply so that the sampling results can fully cover the rapid change process during the transient phase of load connection. During the sampling process, by monitoring the trend of output current changes, it is possible to determine in real time whether there are abnormal change ranges caused by load connection or sudden changes in load status. When the rate of change of the output current exceeds the threshold determined based on the statistical characteristics of steady-state operation, a transient load connection event is identified, and the transient data acquisition window is automatically opened with this moment as the event start point; when the output current changes back to the stable range, the acquisition window is closed. It should be noted that, in order to facilitate horizontal comparison between different load access events, the data of each load access event is aligned to a unified time base, and the event trigger time is mapped to a unified time zero point, so as to ensure that the feature parameters extracted later have consistent physical meaning in the time dimension. S12. Based on the transient output current sequence, the integral form of the current change rate is used to quantify each load connection process, constructing a connection strength index that reflects the degree of connection impact. This transforms the load connection behavior from a single current amplitude description to a behavioral representation of the overall impact characteristics of the connection process, thus more realistically reflecting the transient impact of the connection process on the power supply's internal components. Specifically: Based on the transient output current sequence, a descriptor for the load connection intensity is constructed for each load connection event to reflect the comprehensive transient impact on the rail-mounted power supply during load connection. Its expression is as follows: ; in, This represents the access surge intensity index corresponding to the k-th load access event; This represents the instantaneous value of the output current of the DIN rail power supply. It indicates the rate of change of the output current, used to describe how steep the rise or fall of the current is; This indicates the start time of the triggering of the k-th load connection event; This indicates the moment when the transient change ends and the adjustment and recovery phase begins; S13. Based on the access strength index, further introduce disturbance duration and recovery characteristic indexes to jointly characterize the duration of load access and power recovery capability. This allows for differentiation of differences in time occupation and recovery performance among different load access behaviors, avoiding one-sided judgments based solely on impact magnitude and improving the engineering credibility of behavioral descriptions. Specifically: Based on determining the start and end times of each load access event, a disturbance duration descriptor is constructed to reflect the degree of occupancy of the power dynamic regulation link during the load access process. Its expression is as follows: ; Meanwhile, to characterize the power supply's recovery capability after the load is connected, a recovery characteristic descriptor is introduced, the expression of which is: ; in, This indicates the duration of the k-th load access disturbance; Indicates the disturbance recovery characteristics index; This represents the steady-state output current value after the corresponding load is connected; This indicates the moment when the output current re-enters the steady-state fluctuation range; S14. The access strength, disturbance duration, and recovery characteristics corresponding to the same load access event are uniformly encapsulated to form a structured load access behavior vector. A multi-load access behavior sequence is then constructed in chronological order, transforming discrete and random access events into continuously analyzable measurement results. This provides stable input for subsequent health status inference and adaptive adjustment. Specifically: The access surge intensity index, disturbance duration index, and recovery characteristic index corresponding to the same load access event are uniformly encapsulated to construct a load access behavior vector: ; And in chronological order, multiple load access behavior vectors are combined into a multi-load access behavior sequence: ; in, B represents the load access behavior vector of the k-th load access event; B represents the sequence of multiple load access behaviors; n represents the number of load access behaviors.

[0019] S2. Based on the multi-load access behavior sequence obtained in step S1, the load access impact, duration, and recovery characteristics are converted into equivalent stress inputs. A recursive evolution method is used to infer the internal electrical and thermal stress states of the rail-mounted power supply. After normalization and stability correction, an internal stress state observation sequence that can be directly used for health assessment is formed, realizing quantifiable monitoring of the internal operating state under multi-load access conditions. The specific implementation process is as follows: S21. The impact intensity, disturbance duration, and recovery characteristics in the load access behavior vector obtained in step S1 are weighted and mapped to the equivalent stress input of each load access on the power supply. The weight parameters are calibrated according to the electrical and thermal characteristics of the device, forming a quantitative index that can directly reflect the contribution of internal stress, providing a basic input for subsequent evolution inference, specifically: Based on the load access behavior vector obtained in step S1, the kth load access behavior vector This is mapped to the equivalent stress input generated inside the power supply, constructing the equivalent stress input corresponding to the load connection behavior, that is, transforming externally measurable behavior into a quantitative indicator that can be used to infer the internal state. ; in, This represents the equivalent stress input to the power supply caused by the k-th load connection event; , and This represents the weight coefficient for the behavioral dimension; S22. Based on the equivalent stress input generated by each load connection, a recursive evolution formula is used to update the comprehensive stress state inside the power supply. The stress accumulation caused by the current load connection is superimposed with the previous state. At the same time, the natural recovery capability of the power supply is considered, dynamically reflecting the trend of the internal stress evolving with multiple load connections. This provides continuous state information for subsequent decomposition and health assessment. Specifically: Define the internal stress state of the rail-mounted power supply after the kth load connection as follows: Updated using a recursive relationship: ; in, This represents the internal overall stress state after the k-th load application; This indicates the stress state after the previous load was applied; Represents the stress attenuation coefficient, satisfying ; S23. The overall internal stress state is further decomposed into electrical stress and thermal stress components. By using different attenuation coefficients to handle the impact and sustained effects, the inference results are made more consistent with the internal physical characteristics of the power supply and the response law of the device. This reflects both the peak value of transient electrical stress and the accumulation and recovery of thermal stress, providing an interpretable stress dimension for health assessment. Specifically: Comprehensive stress state Decomposed into electrical stress components and thermal stress components: ; in, This represents the electrical stress component after the k-th load connection; This represents the thermal stress component after the k-th load application; and These correspond to the attenuation coefficients of electrical stress and thermal stress, respectively. This represents the electrical stress component after the (k-1)th load connection; This represents the thermal stress component after the (k-1)th load application; S24. The inferred internal stress state is normalized and stability corrected, the stress changes during each load access cycle are mapped to a comparable range, and the influence of outliers is suppressed through continuous cycle consistency checks, forming a stable and reliable internal stress observation sequence that can be directly used for power supply health status evolution assessment and adaptive control decisions. Specifically: The inferred internal stress state is normalized and stability corrected to construct an observation of the internal stress state: ; in, This represents the normalized internal stress observation values; This indicates the upper limit of the stress reference determined by the power supply design parameters.

[0020] S3. Based on the normalized internal stress observations output in step S2, dynamic assessment and evolution prediction of the health status of the rail-mounted power supply are achieved through initial health quantification, short-term change analysis, long-term trend smoothing, and multi-dimensional risk level determination. This provides a reliable decision-making basis for subsequent adaptive control, taking into account the dual-component characteristics of electrical and thermal stress, and can distinguish between transient impacts and cumulative degradation. This enables quantifiable, continuous, and operable health management. The specific implementation process is as follows: S31. The normalized electrical stress and thermal stress observations output in step S2 are converted into dual-component initial health status indices. Electrical stress health reflects transient shock resistance, and thermal stress health reflects the cumulative heat load level. Through dual-component quantization, interpretable input data is provided for subsequent short-term and long-term health analyses. Specifically: Based on the normalized internal stress observations output in step S2 Define the initial quantity of health status : ; in, This represents the initial health status indicator corresponding to the kth load connection. This represents the electrical stress health index after the k-th load connection; This represents the thermal stress health index after the k-th load connection; and These represent the normalized electrical stress and thermal stress observations, respectively. S32. Calculate short-term changes in health status indicators for continuous load connection, analyze the health increases and decreases of electrical stress and thermal stress components respectively, set thresholds to identify short-term abnormalities or deterioration events, distinguish between instantaneous health declines caused by transient impacts and recoverable fluctuations, provide short-term references for risk warning and subsequent trend analysis, and ensure that health assessments are sensitive to transient fluctuations but do not excessively interfere with long-term judgments. Specifically: Define a vector of short-term health changes: ; And perform anomaly threshold determination: like ; in, Represents a vector of short-term health changes; This indicates the threshold for short-term changes and can be calibrated based on device withstand voltage / thermal capacity experiments. S33. The dual-component health status is smoothed using an exponentially weighted moving average method to form a long-term health status indicator. This eliminates short-term fluctuations and reflects the accumulation of internal stress and the trend of health evolution. By adjusting for the impact of short-term fluctuations through a smoothing coefficient, the long-term health change trajectory can be accurately depicted, providing a stable and reliable data foundation for trend determination and risk level classification. Specifically: The health status of continuous load connections is processed using an exponentially weighted moving average (EWMA) to eliminate short-term fluctuations and generate long-term health status indicators. : ; in, Indicators of long-term health status, with initial conditions as follows: ; Represents the smoothing coefficient, ranging from... It is used to adjust the impact of short-term fluctuations on long-term trends; S34. Calculate health trend indicators by combining long-term health status and short-term changes, and classify the electrical stress and thermal stress components into multi-dimensional risk levels, including green safety, yellow warning, orange risk, and red danger. This directly guides subsequent adaptive control and early warning strategies, forming a closed-loop analysis from internal stress to health status to risk level. Specifically: Based on long-term health status and short-term changes, a comprehensive risk assessment index is constructed, and a multi-dimensional risk level determination is made. Define health trend indicators: ; according to and The combination of these factors categorizes health status into four risk levels (green, safe → red, dangerous), as detailed in Table 1: Table 1. Detailed Classification of Levels ; in, This represents the health trend indicator (long-term change); m represents the historical window length, which is set according to the load connection frequency and power supply cumulative stress characteristics. and These represent the safety and health thresholds for electrical stress and thermal stress, respectively, based on experimental calibration of the device's rated withstand capacity and safety margin. This indicates the health component of the electrical stress that the power supply experiences during long-term operation; It represents the health component of thermal stress that the power supply experiences during long-term operation.

[0021] S4. Using the dual-component long-term health status, health trend, and risk level output in step S3 as core constraints, an adaptive adjustment mechanism is constructed for multi-load dynamic access scenarios. Through health constraint judgment, flexible power adjustment, health closed-loop update, and multi-load access optimization, the risk-controllable operation and lifespan-friendly management of the rail-mounted power supply under multi-load concurrent access conditions are achieved. This transforms the load access behavior from static power allocation to a dynamic closed-loop control process driven by health status. The specific implementation process is as follows: S41. Based on long-term health status, electrical and thermal stress components, and load transient impact characteristics, a health constraint judgment is performed on each load to be connected. An initial connection decision is generated by combining risk level and load priority to determine whether the load can be directly connected, delayed, or enter the power regulation process. This suppresses the sudden impact of high-risk loads on the power supply's health status from the source. Specifically: Let the load to be connected be numbered j, and its transient impact characteristic vector be: (Based on the load access behavior description defined in step S1), define the health constraint judgment function: ; in, This represents the transient access behavior characteristic vector of the j-th load. and These represent the lower safety thresholds for the healthy components of electrical stress and thermal stress, respectively, and are obtained based on the rated life of power supply devices, electrical safety standards, and long-term test calibration. The minimum allowed threshold for load access priority is set by the system operation policy or security policy; This indicates the feasibility determination result for the access of the j-th load. A value of 1 indicates that the load is allowed to access, while a value of 0 indicates that the load needs to be accessed with a delay or reduced power. This indicates an indicator function, which is 1 if the condition is true and 0 otherwise. It should be noted that when the risk level is red, all low-priority loads are delayed in access; at the yellow level, some loads can be accessed and enter the power regulation step; and the loads can be sorted by priority to form an initial access queue. S42. For loads that cannot be connected at full power under health constraints, the proportion of power that can be connected is dynamically calculated based on the health margin and load impact characteristics. This enables continuous adaptive scaling of the load power, allowing the load to participate in operation in a soft-connection mode. This avoids the concentrated accumulation of electrical and thermal stresses while meeting functional requirements. Specifically: When the load cannot be connected at full power due to health constraints, dynamic power adjustment is generated to allow the load to be connected partially while ensuring health and safety. Define adaptive power regulation: ; in, This represents the adaptive access power allocated to the j-th load under health constraints. This represents the rated operating power of the j-th load, which is determined by the technical specifications or nameplate parameters of the load equipment; and These represent the mapping coefficients of load connection impact on the healthy components of electrical stress and thermal stress, respectively, obtained through historical operating data, accelerated life tests, or experimental calibration. S43. After load connection or power regulation is completed, the internal stress and long-term health status are updated in a closed loop based on the actual connected power ratio. This ensures that the health status reflects the real impact of load connection in real time, providing a continuous and traceable health status basis for subsequent load connection decisions. Specifically: After the load is connected or the power is adjusted, the health status needs to be updated in real time for continuous adaptive adjustment. Set update rules: ; ; And recalculate long-term health status: ; in, and These represent the estimated cumulative electrical and thermal stresses within the (k+1)th period, respectively. and These represent the time decay coefficients of electrical stress and thermal stress, respectively, and are calibrated based on the recovery and heat dissipation characteristics of the power supply material. This represents the long-term health state vector of the rail-mounted power supply during the (k+1)th control cycle. S44. In scenarios with multiple concurrent load requests, a dynamic priority score is generated by comprehensively considering health status, load impact characteristics, and power requirements. This optimizes the scheduling of load access order and power allocation, improving overall load access efficiency and system stability while ensuring controllable health risks. Specifically: When multiple loads request access simultaneously, optimized sequence scheduling ensures a safe health status while improving load throughput. Define load balancing priority scoring: ; according to Loads are prioritized and connected sequentially from highest to lowest priority. Lower priority loads are adjusted in real time based on their health status, with delays or power levels updated accordingly. This forms a closed loop; in, This represents the actual power demand requested by the j-th load in the current access request, reported by the load operating status or the control system. , and The weighting coefficients for electrical stress, thermal stress, and power demand in the priority scoring are set by the system operation strategy, load type, and power supply lifetime target.

[0022] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A health adaptive control method for a rail-mounted power supply under multi-load conditions, characterized in that, The specific implementation steps include the following: S1. Real-time acquisition of the output current and output voltage of the rail-mounted power supply under multi-load random access conditions, identification of load access transient events, extraction and construction of a structured load access behavior vector reflecting the access impact intensity, disturbance duration and recovery characteristics based on the transient sequence of output current, and formation of a multi-load access behavior sequence in chronological order. S2. The load access behavior vector in the multi-load access behavior sequence is mapped to the equivalent stress input through weighted combination. The internal comprehensive stress state of the rail-mounted power supply is dynamically updated and inferred using a recursive evolution method. The internal comprehensive stress state is decomposed into electrical stress components and thermal stress components. After normalization, an internal stress state observation sequence is formed. S3. Based on the internal stress state observation sequence, the initial health status index with two components is calculated. The initial health status index is subjected to short-term change analysis and long-term trend smoothing to obtain the long-term health status index. The multi-dimensional risk level is then divided in combination with the health trend index. S4. Using long-term health status indicators, health trend indicators, and multi-dimensional risk levels as constraints, the system performs health constraint judgment and dynamic priority scoring on the loads applying for access. For loads that are determined to be allowed to access, the system performs flexible power adjustment based on the health margin. After the load access is completed, the system performs closed-loop updates on the internal stress state and health state to achieve dynamic optimization scheduling of multiple load access.

2. The health adaptive control method for a rail-mounted power supply under multi-load conditions according to claim 1, characterized in that, In step S1, the specific process of constructing the load access behavior vector includes: The start time of the load connection transient event is determined by using the rate of change of the output current exceeding a threshold determined based on the statistical characteristics of steady-state operation. The event ends when the output current changes back to the steady-state range. Based on the transient sequence of the output current, the rate of change of the current from the start time to the end time is integrated to calculate the index characterizing the intensity of the access impact. An index characterizing the duration of the disturbance is calculated based on the start and end times; Based on the time required for the output current to reach a steady-state value after the load connection event ends, an index characterizing the recovery characteristics is calculated. The access impact intensity index, disturbance duration index, and recovery characteristic index calculated for the same load access event are combined and encapsulated to form the load access behavior vector for that event.

3. The health adaptive control method for a rail-mounted power supply under multi-load conditions according to claim 2, characterized in that, In step S2, the specific process of the recursive evolution method is as follows: The three dimensions of the load access behavior vector—access impact intensity, disturbance duration, and recovery characteristics—are weighted and combined according to the weighting coefficients calibrated based on the electrical and thermal characteristics of the power supply devices, and mapped to the equivalent stress input corresponding to the load access for the kth time. The internal comprehensive stress state after the kth load connection is obtained by adding the equivalent stress input of the kth load connection and the internal comprehensive stress state after the (k-1)th load connection by the stress attenuation coefficient. The internal stress state can be further decomposed into electrical stress components and thermal stress components: The electrical stress component is obtained by adding the current equivalent stress input's impact intensity portion to the previous electrical stress component multiplied by the electrical stress attenuation coefficient; The thermal stress component is obtained by adding another part of the current equivalent stress input to the value of the previous thermal stress component multiplied by the thermal stress attenuation coefficient.

4. The health adaptive control method for a rail-mounted power supply under multi-load conditions according to claim 3, characterized in that, In step S3, the specific calculation process for the initial health status index of the two components is as follows: The normalized electrical stress observation values ​​output in step S2 are converted into electrical stress health indicators through the first preset mapping relationship. The normalized thermal stress observation values ​​are converted into thermal stress health indicators through a second preset mapping relationship; The electrical stress health index and the thermal stress health index together constitute a two-component initial health state index that reflects the power supply's ability to withstand transient impacts and the level of cumulative thermal load.

5. The health adaptive control method for a rail-mounted power supply under multi-load conditions according to claim 4, characterized in that, In step S3, the long-term trend smoothing process specifically employs the exponentially weighted moving average method, as follows: An exponentially weighted moving average is applied to the health status of continuous load connections to generate a long-term health status index. : ; in, Indicators of long-term health status, with initial conditions as follows: ; Represents the smoothing coefficient, ranging from... It is used to adjust the impact of short-term fluctuations on long-term trends; This represents the initial health status indicator corresponding to the kth load connection.

6. The health adaptive control method for a rail-mounted power supply under multi-load conditions according to claim 5, characterized in that, Step S3, the specific process of classifying multidimensional risk levels includes: The electrical stress health component and thermal stress health component in the long-term health status index are compared with the electrical stress safety health threshold, thermal stress safety health threshold, electrical stress warning threshold, thermal stress warning threshold, electrical stress danger threshold and thermal stress danger threshold calibrated based on the rated withstand capacity and safety margin of the device. Calculate the health trend index, which represents the amount of change in long-term health status within a set historical window. Based on the threshold range of the electrical stress health component and the thermal stress health component, as well as whether the health trend indicator is declining, the risk level is comprehensively determined as green safety level, yellow warning level, orange risk level, and red danger level.

7. The health adaptive control method for a rail-mounted power supply under multi-load conditions according to claim 6, characterized in that, In step S4, the specific process for determining health constraints is as follows: Obtain the transient impact characteristic vector of the load to be connected. This vector includes the estimated access impact intensity, disturbance duration, and recovery characteristics when the load is connected. Read the electrical stress health component and thermal stress health component from the long-term health status indicators at the current moment; Determine whether the electrical stress health component is higher than the electrical stress safety lower limit threshold, whether the thermal stress health component is higher than the thermal stress safety lower limit threshold, and whether the load's preset access priority is higher than the set minimum allowable threshold. The health constraint decision is to allow access only if all judgment conditions are met; otherwise, the decision is to delay access and enter the power regulation process.

8. The health adaptive control method for a rail-mounted power supply under multi-load conditions according to claim 7, characterized in that, In step S4, the specific process of power flexible adjustment is as follows: When the load health constraint determination result is that it needs to enter the power regulation process, the electrical stress health margin is determined based on the difference between the current electrical stress health component and the electrical stress safety lower limit threshold, and the thermal stress health margin is determined based on the difference between the current thermal stress health component and the thermal stress safety lower limit threshold. By combining the mapping coefficients of electrical stress and thermal stress in the transient impact characteristic vector of the load, the adaptive access power is dynamically calculated. The adaptive access power is less than the rated operating power of the load, and its value is positively correlated with the electrical stress health margin and the thermal stress health margin.

9. A health adaptive control method for a rail-mounted power supply under multi-load conditions according to claim 8, characterized in that, In step S4, the specific process of closed-loop update is as follows: After the load is actually connected according to the rated power and adaptive connection power, the actual contribution of this connection to electrical stress and thermal stress is calculated based on the ratio of the actual connection power to the rated power. The actual contribution is superimposed onto the electrical stress component and thermal stress component before the update, and then multiplied by the corresponding electrical stress time decay coefficient and thermal stress time decay coefficient respectively to obtain the updated electrical stress component and thermal stress component. Based on the updated electrical stress and thermal stress components, normalization and health status index calculations are performed again to complete the closed-loop update of long-term health status indicators.

10. A health adaptive control method for a rail-mounted power supply under multi-load conditions according to claim 9, characterized in that, In step S4, the specific process of generating dynamic priority scores is as follows: When several loads simultaneously request access, a dynamic priority score is calculated for each load. The calculation of this score takes into account the correlation between the current power supply’s long-term electrical stress health level and the load’s estimated electrical shock characteristics, the correlation between the long-term thermal stress health level and the load’s estimated thermal shock characteristics, and the magnitude of the load’s own power demand. Among them, the estimated electrical and thermal shock characteristics of the load have a negative impact on its score; During the calculation, weight coefficients determined by the system operation strategy are assigned to the electrical stress health correlation item, thermal stress health correlation item, and power demand item respectively; The loads applying for access are sorted according to the calculated dynamic priority score. Loads with higher scores are given priority to access permission, while loads with lower scores are delayed in access or enter the power adjustment process based on the real-time updated health status, thus forming a closed-loop dynamic access scheduling.

Images