Intelligent operation monitoring method and system for power conversion cabinet

Abstract

The invention discloses an intelligent operation monitoring method and system for a battery changing cabinet, and the method comprises the steps: obtaining a thermoelectric imbalance index through standardized deviation calculation and exponential smoothing based on the 24-hour sliding statistical data of the voltage and temperature of a single battery compartment in the battery changing cabinet; constructing a multi-dimensional topological node set based on multi-physical information of a thermoelectric imbalance index over-threshold bin, and calculating a thermoelectric coupling spreading potential in combination with a minimum spanning tree structure; based on the thermoelectric coupling spreading potential and the waiting queue length, integrating index adjustment of the time period adaptive weight and tail risk penalty to obtain a service toughness index; obtaining an available safety margin through dynamic redundancy compression and safety lower limit protection based on the service toughness index and the current idle warehouse number; and based on the difference between the available safety margin of the battery changing cabinet and the service toughness index of the adjacent cabinet, obtaining the regional collaborative guidance intensity through standard deviation judgment and index attenuation mapping. According to the invention, the intelligent level of the battery changing cabinet in the aspects of safety risk perception and service capability adjustment is improved.

Description

Technical Field

[0001] This invention relates to the field of electrical engineering, and in particular to a method and system for intelligent operation monitoring of a battery swapping cabinet. Background Technology

[0002] Current battery swapping cabinet monitoring generally suffers from three major bottlenecks: lagging safety awareness, a disconnect between scheduling strategies and hardware risks, and isolated cabinet operation lacking coordination. Traditional methods rely on static threshold alarms, making it difficult to identify sub-health conditions such as thermal-electric imbalance in the early stages; service allocation does not consider safety factors such as aging and heat dissipation, easily leading to the introduction of fully charged batteries in high-risk compartments; and each cabinet operates independently, failing to dynamically guide user allocation based on regional status, resulting in both localized overload and resource idleness. Based on this, this invention proposes an intelligent operation monitoring method and system for battery swapping cabinets. Summary of the Invention

[0003] This invention provides a method for intelligent operation monitoring of a battery swapping cabinet, characterized by comprising: S10. Based on the 24-hour sliding statistical data of voltage and temperature of a single battery compartment in the battery swapping cabinet, a thermoelectric imbalance index is obtained through standardized deviation calculation and exponential smoothing, which is used to characterize the thermal and electrical deviation of the battery in the compartment. S20. Based on the multi-physical information of the warehouse with the thermoelectric imbalance index exceeding the threshold, a multi-dimensional topological node set is constructed. Combined with the minimum spanning tree structure, the thermoelectric coupling propagation potential reflecting the spatial aggregation risk of high-risk warehouses is calculated. S30. Based on thermoelectric coupling propagation potential and waiting queue length, the service resilience index that integrates hardware risk and service pressure is obtained by combining time period adaptive weighting with tail risk penalty. S40. Based on the service resilience index and the current number of idle warehouses, after dynamic redundancy compression and safety lower limit protection, the available safety margin that determines the upper limit of the number of serviceable warehouses is obtained. S50: Based on the difference between the available safety margin of the battery swapping cabinet and the service resilience index of neighboring cabinets, the regional collaborative guidance strength driving user diversion is obtained through standard deviation discrimination and exponential decay mapping, thereby achieving load balancing and risk dispersion without a central coordination mechanism.

[0004] The intelligent operation monitoring method for battery swapping cabinets described above includes the following sub-steps: Based on 24-hour sliding statistical data of voltage and temperature in a single battery compartment within the cabinet, a thermoelectric imbalance index is obtained through standardized deviation calculation and exponential smoothing. This index is used to characterize the thermal and electrical deviations of the batteries in that compartment. Real-time acquisition of battery compartment voltage and temperature data, and calculation of statistics and sequence sample entropy through exponential decay weighted sliding window, providing a time-sensitive calculation benchmark for thermoelectric imbalance assessment; By standardizing voltage and temperature deviations, introducing dynamic weighting of sample entropy and fusing exponentially smoothed imbalance deterioration rates, an adaptive thermoelectric imbalance index is constructed to quantify the multidimensional anomaly risks of the battery compartment. Through triple verification of dynamic threshold, deterioration trend, and duration, the thermoelectric imbalance index value is transformed into a reliable sub-healthy ward marker and incorporated into cohort management to support upper-level risk decision-making.

[0005] The intelligent operation monitoring method for battery swapping cabinets described above involves converting the thermoelectric imbalance index value into a reliable sub-healthy status marker through triple verification of dynamic thresholds, deterioration trends, and duration. This marker is then incorporated into queue management to support upper-level risk decision-making. The method comprises the following sub-steps: The dynamic threshold mechanism first sets an exponential base threshold, and then superimposes a periodic correction term on the base threshold to simulate the impact of diurnal temperature variation on normal fluctuations. If the current thermoelectric imbalance index is greater than the correction term, the battery compartment is marked as a suspected anomaly, and a trend analysis mechanism is used to verify whether it shows a continuous deterioration trend. Only when the thermoelectric imbalance index is greater than the correction term and shows a continuous deterioration trend will the battery compartment be upgraded to a candidate for sub-health. The duration verification mechanism implements time window cumulative verification for sub-health candidate battery compartments. If the thermoelectric imbalance index of the compartment exceeds the limit for more than 30 minutes, the battery compartment is officially marked as sub-healthy and added to the priority monitoring queue. If the thermoelectric imbalance index value falls below the correction item or does not show a deteriorating trend, the battery compartment returns to a healthy state.

[0006] The intelligent operation monitoring method for battery swapping cabinets described above involves constructing a multi-dimensional topological node set based on the multi-physical information of the compartment with a thermo-electric imbalance index exceeding the threshold, and calculating the thermo-electric coupling propagation potential reflecting the spatial clustering risk of high-risk compartments using a minimum spanning tree structure. This method is specifically divided into the following sub-steps: High-risk battery compartments are selected from sub-healthy compartments, and their spatial coordinates, thermoelectric parameters, internal resistance and cooling status are integrated to construct a multi-dimensional topological node set, providing structured risk source data for thermal spread calculation; A complete graph is constructed based on the physical coordinates of the high-risk battery compartments, and a minimum spanning tree is generated using the Prim algorithm to depict the spatial distribution and adjacency relationships of the abnormal compartments in a compact manner. The potential risk of thermal runaway chain reaction in high-risk warehouses is quantified by calculating the thermoelectric coupling propagation potential based on the minimum spanning tree structure and multidimensional topological node set.

[0007] The above-described intelligent operation monitoring method for battery swapping cabinets, which integrates thermoelectric coupling propagation potential and waiting queue length, incorporates time-period adaptive weighting and tail risk penalty to obtain a service resilience index that combines hardware risk and service pressure. This method comprises the following sub-steps: The instantaneous load density is calculated by the current waiting queue length, and the normalized load index is calculated by applying a nonlinear compression function based on the load density, providing a basis for scheduling strategies. The load weight and security weight for the current moment are dynamically generated based on historical load data and normalized to achieve adaptive security and service co-optimization under different loads and time periods. The service resilience index is calculated based on thermoelectric coupling propagation potential, normalized load index, waiting queue length, and dual adaptive weighting.

[0008] The intelligent operation monitoring method for battery swapping cabinets described above, based on the service resilience index and the current number of idle bays, obtains the available safety margin that determines the upper limit of the number of serviceable bays through dynamic redundancy compression and safety lower limit protection. This method specifically comprises the following sub-steps: Set a basic number of redundant warehouses and an absolute bottom line, and dynamically adjust the redundancy requirements according to the ambient temperature, transforming physical safety constraints into adaptively adjustable scheduling boundary conditions; The available service ratio of the battery swapping cabinet is dynamically adjusted based on the service resilience index to achieve risk-adaptive elastic resource scheduling and maximize service efficiency while ensuring safety. Dual security protection is implemented based on the service resilience index, and the proportion of available services is converted into the number of executable scheduling warehouses that balance security and efficiency based on the service resilience index.

[0009] The above-described intelligent operation monitoring method for battery swapping cabinets, based on the difference between the available safety margin of the battery swapping cabinet and the service resilience index of neighboring cabinets, obtains the regional collaborative guidance strength driving user diversion through standard deviation discrimination and exponential decay mapping, thereby achieving load balancing and risk dispersion without a central coordination mechanism. Specifically, it consists of the following sub-steps: The battery swapping cabinet communicates with neighboring cabinets through periodic encrypted broadcasts, combined with smooth processing via sliding windows, to build a lightweight regional resilient state awareness network in a decentralized architecture. By introducing the distance-weighted service resilience index standard deviation, the resilience difference between the battery swapping station and its neighboring stations is quantified. When the difference is significant and the battery swapping station has the highest risk, collaborative scheduling is triggered. After confirming the regional imbalance, based on the difference in service resilience index between the battery swapping cabinet and its neighboring cabinets and its own availability margin, a collaborative guidance strength is generated through nonlinear mapping, and a graded prompt is triggered accordingly to achieve adaptive user diversion driven by risk differences.

[0010] This invention also provides an intelligent operation monitoring system for battery swapping cabinets, comprising: Thermoelectric Imbalance Module: Based on 24-hour sliding statistical data of voltage and temperature of a single battery compartment in the battery swapping cabinet, a thermoelectric imbalance index is obtained through standardized deviation calculation and exponential smoothing, which is used to characterize the thermal and electrical deviation of the battery in the compartment. Thermoelectric Coupling Module: Based on the multi-physical information of the thermoelectric imbalance index exceeding the threshold, a multi-dimensional topological node set is constructed. Combined with the minimum spanning tree structure, the thermoelectric coupling propagation potential reflecting the spatial aggregation risk of high-risk warehouses is calculated. Service resilience module: Based on thermoelectric coupling propagation potential and waiting queue length, it integrates time-period adaptive weighting of exponential adjustment and tail risk penalty to obtain a service resilience index that combines hardware risk and service pressure. Safety margin module: Based on the service resilience index and the current number of idle warehouses, through dynamic redundancy compression and safety lower limit protection, the available safety margin that determines the upper limit of the number of warehouses that can be served is obtained; Collaborative guidance module: Based on the difference between the available safety margin of the battery swapping cabinet and the service resilience index of neighboring cabinets, the regional collaborative guidance intensity driving user diversion is obtained through standard deviation discrimination and exponential decay mapping, thereby achieving load balancing and risk dispersion without a central coordination.

[0011] The beneficial effects achieved by this invention are as follows: This invention improves the intelligence level of the battery swapping cabinet in three aspects: safety risk perception, adaptive adjustment of service capabilities, and decentralized collaborative scheduling, realizing a leap from passive alarm to active resilience control. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0013] Figure 1 This is a flowchart of an intelligent operation monitoring method for a battery swapping cabinet provided in Embodiment 1 of this application.

[0014] Figure 2 This is a schematic diagram of an intelligent operation monitoring system for a battery swapping cabinet provided in Embodiment 2 of this application. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1 like Figure 1 As shown, Embodiment 1 of this application provides a method for intelligent operation monitoring of a battery swapping cabinet, including: S10. Based on the 24-hour sliding statistical data of voltage and temperature of a single battery compartment in the battery swapping cabinet, a thermoelectric imbalance index is obtained through standardized deviation calculation and exponential smoothing, which is used to characterize the thermal and electrical deviation of the battery in the compartment.

[0017] S11. Real-time acquisition of battery compartment voltage and temperature data, calculation of statistics and sequence sample entropy through exponential decay weighted sliding window, providing a time-sensitive calculation benchmark for thermoelectric imbalance assessment.

[0018] The terminal voltage and surface temperature data of each battery compartment in the battery swapping cabinet are acquired in real time with a fixed sampling period of 10 seconds. Independent 24-hour sliding window data structures for voltage and temperature are constructed. The window adopts an exponential decay weighting mechanism and sets the basic weight of the latest data point. The weight of historical data points is 0.95, calculated using the formula... The calculation generates a weighted sequence that decays with each time step, where k represents the time step number moving backward from the current moment. k gradually increases from 1, with k=1 representing the latest point. This mechanism effectively balances short-term fluctuation response and long-term trend capture. Based on the weighted window, four core statistics are calculated in parallel: voltage weighted mean, voltage weighted standard deviation, temperature weighted mean, and temperature weighted standard deviation. The voltage weighted mean is obtained by multiplying the voltage value at each moment within the sliding window by its corresponding exponentially decaying weight, summing the results, and then dividing by the sum of all weights. The temperature weighted mean is calculated in the same way. The voltage weighted standard deviation is obtained by first calculating the square of the difference between each voltage value and the voltage weighted mean, then multiplying each difference by its corresponding weight, summing the results, dividing by the total weight, and finally taking the square root. The temperature weighted standard deviation is calculated in the same way. Simultaneously, the sample entropy of the voltage and temperature sequences is calculated using a sample entropy algorithm. The smaller the entropy value, the more regular the temperature and voltage fluctuations; the larger the value, the more random and disordered the fluctuations. All statistical calculations are implemented using an incremental update algorithm. Only when a new data point arrives is the current statistical value quickly updated using the cumulative sum and weights from the previous time step, avoiding reprocessing the entire window of data each time. The incremental update algorithm uses single-precision floating-point arithmetic, and historical data is stored in a circular buffer. Each battery compartment retains only the most recent 72 hours of raw data for recalculation and verification. During normal operation, memory usage is strictly controlled to 16KB per compartment.

[0019] S12. By standardizing voltage and temperature deviations, introducing dynamic weighting of sample entropy and integrating exponentially smoothed imbalance deterioration rates, an adaptive thermoelectric imbalance index is constructed to quantify the multidimensional abnormal risks of the battery compartment.

[0020] Based on statistical characteristics and sequence sample entropy, a thermoelectric imbalance index is calculated to identify in real-time and quantitatively whether a single battery compartment simultaneously exhibits coupled risk states of abnormal thermal management and deviations in electrochemical behavior, thereby providing accurate early warnings in the early stages of thermal runaway or performance degradation. The specific formula is as follows: , Let be the thermoelectric imbalance index of the i-th battery compartment. The standardized deviation of the voltage of the i-th battery compartment is the voltage value of the i-th battery compartment at the current time t. The result is obtained by subtracting the voltage-weighted mean within its sliding window and then dividing by the corresponding voltage-weighted standard deviation. Let be the absolute value of the standardized deviation of the voltage of the i-th battery compartment. This is the entropy weighting coefficient, controlling the strength of the sample entropy's adjustment of the current deviation. The initial value is 0.5, and the range is [0.3, 0.6]. It is adjusted according to the false alarm rate: if the thermoelectric imbalance index fluctuates highly and there are frequent false alarms, the value should be appropriately increased. If early anomalies are missed, the risk should be appropriately reduced. , Let be the sample entropy of the voltage sequence of the i-th battery compartment, used to quantify the orderliness of historical fluctuations. Temperature standardization bias is the surface temperature of the i-th battery compartment at the current time t. The result is obtained by subtracting the temperature-weighted mean within the sliding window and then dividing by the corresponding temperature-weighted standard deviation. Let be the absolute value of the standardized deviation of the temperature of the i-th battery compartment. Let be the sample entropy of the i-th battery compartment temperature sequence. To mitigate the rate sensitivity coefficient, the imbalance rate of change term is determined. The intensity of the adjustment to the denominator is initially 3, and the range is [2,5]. To comprehensively measure thermoelectric loss, This represents the derivative of the overall thermoelectric imbalance with respect to time t, i.e., calculating its instantaneous rate of change. A positive value indicates that the imbalance is worsening, while a negative value indicates that it is tending to recover. This indicates that the derivative result is smoothed by applying an exponential moving average (EMA). EMA is a lightweight low-pass filtering method used to suppress high-frequency noise caused by numerical differentiation. The smoothing coefficient controls the smoothing effect on the rate of change of imbalance. The initial value is 0.8, and the range is [0.7, 1). The closer the EMA is to 1, the smoother the response, the greater the hysteresis, the stronger the noise immunity but the slower the response. The smaller the value, the faster the response, but the more susceptible it is to sampling jitter.

[0021] S13. Through triple verification of dynamic threshold, deterioration trend and duration, the thermoelectric imbalance index value is transformed into a reliable sub-healthy warehouse marker and incorporated into the queue management to support upper-level risk decision-making.

[0022] The thermoelectric imbalance index achieves highly robust identification of the sub-health state of the battery compartment through a three-layer mechanism of dynamic threshold, trend analysis, and duration verification. The dynamic threshold mechanism sets the basic threshold of the index to... Based on the threshold, a periodic correction term is superimposed. The simulation of diurnal temperature variations on normal fluctuations further increases the threshold during high-temperature environments or periods of high load to avoid misjudgments caused by reasonable temperature rises. The specific formula is as follows: , This represents the daily average baseline threshold offset, reflecting the average risk level throughout the day. The coefficients of the cosine and sine terms, respectively, together determine the amplitude and phase of the periodic oscillation. The process of obtaining the value is as follows: By collecting 7 days of historical thermoelectric imbalance index data, the hourly average thermoelectric imbalance index is calculated by aggregating the data. Then, the least squares method is used to fit a first-order Fourier model containing constant, cosine, and sine terms to the 24 mean points, and the solution is obtained. . Let π be the mathematical constant pi, and t be the current hour, expressed in 24-hour format. This is an uncertainty compensation term that enables the threshold to adapt to the severity of load fluctuations over different time periods. Let be the confidence coefficient, and take . Corresponding to the 95% confidence interval It can be adjusted according to the false alarm tolerance. Let be the standard deviation of the historical thermoelectric imbalance index at hour t, calculated based on 7 days of historical data for the same hour. If the current... If the condition is not met, the warehouse is marked as a suspected anomaly, and its continued deterioration trend is verified. The trend analysis mechanism utilizes the imbalance rate of change term. If the derivative term is greater than zero for three consecutive sampling points and all exceed the preset slope threshold, then the battery compartment condition is determined to be deteriorating. If the derivative term is high but negative or close to zero, it is considered transient interference such as voltage fluctuations or communication jitter, and no alarm is triggered. Only when... Only when the derivative term remains consistently positive is the battery compartment upgraded to a sub-health candidate. A duration verification mechanism performs time window cumulative verification on battery compartments in the sub-health candidate state: a timer is started to record the battery compartment's condition. If the battery compartment continues to exceed the limit for more than 30 minutes, it will be officially marked as sub-healthy and added to the priority monitoring queue. If it continues beyond this period... If the value falls below the threshold or the derivative term becomes negative, the timer is cleared and the state is reset back to a healthy state.

[0023] S20. Based on the multi-physical information of the warehouse with the thermoelectric imbalance index exceeding the threshold, construct a multi-dimensional topological node set, and combine it with the minimum spanning tree structure to calculate the thermoelectric coupling propagation potential that reflects the spatial aggregation risk of high-risk warehouses.

[0024] S21. Select high-risk battery compartments from the sub-healthy compartments, integrate their spatial coordinates, thermoelectric parameters, internal resistance and cooling status, and construct a multi-dimensional topological node set to provide structured risk source data for thermal propagation calculation.

[0025] From the output sub-healthy battery monitoring queue, high-risk battery compartments with a thermoelectric imbalance index greater than 0.7 were further screened as potential thermoelectric risk sources. For each high-risk compartment, its physical coordinates within the battery cabinet were extracted to construct a spatial topology, and its key real-time status parameters were simultaneously collected, including surface temperature, terminal voltage, and internal resistance estimated based on voltage-current transient response. Simultaneously, the operating status of the cabinet's cooling system was read, specifically the fan control signal of the corresponding compartment to characterize local heat dissipation capacity (1 indicates on, 0 indicates off). All the above information, including spatial location, thermoelectric state, internal resistance characteristics, and cooling conditions, was encapsulated into a structured topology node. The topology nodes of all high-risk compartments constitute a multi-dimensional topology node set, completing the transformation from discrete anomaly indicators to risk entities with multi-dimensional physical attributes and spatial relationships.

[0026] S22. Construct a complete graph based on the physical coordinates of the high-risk battery compartments, and generate a minimum spanning tree using the Prim algorithm to depict the spatial distribution and adjacency relationships of the abnormal compartments in a compact manner.

[0027] Taking all the selected high-risk battery compartments as objects, their physical coordinates are used as nodes to construct a complete graph G = (J, E), where the node set J corresponds to the coordinates of each high-risk compartment, and the edge set E contains the connections between all node pairs. The weight of each edge is defined as the Euclidean distance between the two compartments. Then, Prim's algorithm is used to calculate the minimum spanning tree from this complete graph. This algorithm selects any node in the graph as the starting point and adds it to the tree being constructed. Among all edges connecting nodes already added to the tree to nodes not yet added, it selects the edge with the smallest weight and adds the corresponding node not yet added to the tree. This process is repeated iteratively, selecting the edge with the smallest weight connected to the current tree, gradually expanding until all abnormal compartment nodes are covered. Finally, a tree structure with the smallest total edge length and no cycles is generated, connecting all high-risk compartments. The edge set S of the resulting minimum spanning tree not only preserves the spatial adjacency relationships between high-risk compartments but also describes their distribution within the compartments in the most compact way. Through this process, the originally discrete spatial coordinates are transformed into a graph structure with clear geometric meaning and computable properties.

[0028] S23. Calculate the thermoelectric coupling propagation potential based on the minimum spanning tree structure and multidimensional topological node set to quantify the potential risk of high-risk inter-warehouse thermal runaway chain reaction.

[0029] Based on the spatial connectivity of high-risk battery compartments defined by the minimum spanning tree, and by integrating multi-physics information such as temperature, voltage, internal resistance, and active cooling effects from multi-dimensional topological nodes, the potential risk of thermoelectric coupling propagation potential cascading failures between abnormal battery compartments is calculated. The specific formula is as follows: MPCP represents the thermoelectric coupling propagation potential, ranging from [0,1]. A value closer to 0 indicates that high-risk compartments are not only spatially adjacent but also exhibit strong synergistic effects of heat conduction, electrical interference, and aging, drastically increasing the likelihood of thermal runaway propagation. This provides a physically interpretable risk potential measure for cabinet-level safety decisions. S is the minimum spanning tree edge set, where u and v represent the indices of two high-risk battery compartments, i.e., the two endpoint nodes connected by an edge in the minimum spanning tree. (u, v) represents a pair of directly connected high-risk compartments in the minimum spanning tree. Let be the Euclidean distance between _u_ and _v_. This is a thermal conduction coupling term. The greater the temperature difference and the shorter the distance, the stronger the thermal crosstalk represented by this term. The thermal coupling sensitivity coefficient has an initial value of 1.2 and a range of [0.8, 1.5]. The larger the value, the stronger the suppression of thermal crosstalk under the same temperature difference and distance. The temperature difference between the surfaces of bins u and v. This represents the absolute value of the temperature difference. For electrical crosstalk coupling, the electromagnetic compatibility interference risk caused by voltage differences between adjacent compartments is quantified. is the electrical coupling sensitivity coefficient, with an initial value of 4.5 and a range of [3, 6]. The voltage difference between terminals u and v represents the intensity of electrical interference. This is the absolute value of the terminal voltage difference. This is the system nominal voltage, used to normalize the voltage difference. This is the aging synergistic effect term. The aging coupling sensitivity coefficient has an initial value of 1.8 and a range of [1, 2.5]. Let u and v be the estimated internal resistances of the bins, respectively. The average internal resistance of the new battery. As an active cooling suppression term, the effect of air cooling on preventing heat spread is quantified. The cooling suppression coefficient measures the ability of air cooling to reduce thermal crosstalk, and is set to a value of 0.5. These represent the cooling fan status for the hoppers U and V, respectively. A value of 1 indicates the fan is on, and a value of 0 indicates it is off. M represents the maximum Euclidean distance between the two battery compartments inside the cabinet, and M represents the total number of high-risk compartments. The value should be an extremely small positive number to prevent the denominator from being zero. At high-density urban battery swapping stations, the tolerance for missed reports is extremely low, and the value can be appropriately reduced. To enhance sensitivity, especially at remote sites where the cost of false alarms is high, a higher value could be appropriately increased. To enhance robustness, the intelligent operation monitoring method for battery swapping cabinets proposed in this invention was deployed. 30 days of operational data from the battery swapping cabinets were collected, with MPCP < 0.3 and subsequent failures occurring within the next 30 minutes considered as positive samples. A Bayesian optimization algorithm was then used to automatically find the optimal prediction performance. combination.

[0030] S30. Based on the thermoelectric coupling propagation potential and the waiting queue length, an exponential adjustment with time-period adaptive weights and tail risk penalty are combined to obtain a service resilience index that integrates hardware risk and service pressure.

[0031] S31. Calculate the instantaneous load density using the current waiting queue length, and calculate the normalized load index using a nonlinear compression function based on the load density, providing a basis for scheduling strategies.

[0032] First, obtain the length of the waiting queue for user battery swapping requests at the current moment. And in conjunction with the total number of battery swapping compartments N, the instantaneous load density is calculated. This is used to characterize the current service resource occupancy pressure. Subsequently, a normalized load index is obtained by applying the Sigmoid nonlinear compression function to the instantaneous load density, with the specific formula being: The value range is (0,1), where the threshold is... This represents the critical point at which user perception significantly deteriorates, i.e., 60% of the warehouse space is occupied or has queues. e is the base of the natural logarithm, and the sensitivity coefficient. The steepness of the control curve is adjusted to ensure performance under low load. A gradual increase in the load metric output during a given period indicates that users experience no noticeable difference, whereas a high load... The interval load metric output rapidly approaches 1, reflecting a sharp increase in the risk of user churn. The normalized load metric incorporates real-time queuing status and accurately characterizes the non-linear sensitivity of user behavior to service latency through non-linear mapping, providing interpretable and comparable service pressure input for subsequent resilient scheduling.

[0033] S32. Dynamically generate the current load weight and security weight based on historical load data, and perform normalization processing to achieve adaptive security and service collaborative optimization under different loads and time periods.

[0034] Based on a historical load database accumulated over a long period, dynamic decision-making weights for both operations and security are generated for the current time t: First, the load weights are calculated. ,in The historical average load rate for the same period as the current time t reflects the typical level of business activity. A weekday indicator, set to 1 for weekdays and 0 for weekends, is used to capture cyclical patterns. This is the basic load weight bias, which is the minimum load attention when the historical load rate is 0 and it is a non-working day. It reflects the minimum service priority that is still retained in the most idle scenario, with an initial value of 0.4. Historical load sensitivity coefficient, representing the historical average load rate. For every increase of 1 (i.e., 100% load), the base load weight increment controls the amplification effect of business busyness on service priority; the initial value is 0.6. This is the weekday enhancement factor, representing the additional load weight on weekdays compared to weekends under the same time period load. It is used to capture periodic business patterns such as commuting and logistics, with an initial value of 0.2. A safety weight is also calculated. Its design logic complements the load weighting; when the historical load is low, The increase in security weight automatically elevates the alert level, reflecting a strategy that prioritizes security during periods of low business activity. The basic security weight bias represents the minimum level of security concern that is maintained even when the historical load rate reaches its maximum value of 1 and the load is at full capacity. The initial value is 0.3. The low-load safety enhancement factor controls the increment of the safety weight for every 1 unit decrease in historical load, with an initial value of 0.6.

[0035] The load weight and security weight are normalized to obtain the normalized load weight. and normalized safety weights This ensures that the sum of the two is 1 and that they are comparable. The process encodes the spatiotemporal load patterns accumulated over long-term operations, including weekday peaks and nighttime off-peaks, into quantifiable decision parameters, enabling subsequent scheduling strategies to automatically balance the priorities of service efficiency and operational safety according to different time periods.

[0036] S33. Calculate the service resilience index based on thermoelectric coupling propagation potential, normalized load index, waiting queue length, and dual adaptive weights.

[0037] The service resilience index is used to comprehensively assess the overall resilience of a single battery swapping station under current conditions in terms of resisting thermal runaway risks, maintaining service continuity, and coping with equipment aging. The specific formula is as follows: ,in The service resilience index indicates that the higher the value, the greater the probability that the battery swapping cabinet will avoid both safety incidents and service interruptions within the next 30 minutes from the current time t. As a core input item for security risks, The potential for thermoelectric coupling propagation between the battery compartments of a single battery swapping cabinet. The normalized safety weight serves as a dynamic adjustment index in this item; the higher the risk period, the greater the impact of the safety dimension on overall resilience. This indicates the impact of the current service availability of the battery swapping station on the overall resilience of the station. For normalized load metrics, For normalized load weights, Quantifying the impact of the current cooling system's effective operating capacity on overall resilience, a higher real-time heat spread risk and insufficient cooling significantly lower the service resilience index value. The available power for the battery swapping cabinet cooling system. The rated power of the battery swapping cabinet cooling system. For thermal management weight, The base coefficient is 1, with a value range of [0.6, 1.4]. The impact of the overall aging state of the quantified battery pack on its overall toughness. It is the average normalized internal resistance of all battery compartments in the battery cabinet, where N is the total number of compartments in the battery swapping cabinet. Let be the current internal resistance of the i-th battery in the battery swapping cabinet. Internal resistance at the end of battery life For the remaining health margin of the battery, This is a dynamic weighting for the aging dimension. This weighting prevents excessive focus on aging when the battery is new, but actively amplifies the negative impact on resilience once aging becomes apparent. This is the aging baseline coefficient, with a value of 1. This represents the total weight. This is a tail risk penalty term, where Q is the length of the waiting queue. This is the critical queuing threshold. The joint control measures the strength and speed of the battery swapping cabinet resilience index being nonlinearly suppressed when the user waiting queue length exceeds a critical threshold. This is the penalty intensity coefficient, initially set to 2. The nonlinear kurtosis index controls the degree to which the penalty accelerates as the degree of exceeding the limit increases. The initial value is 1.5, and the value range is [1,2].

[0038] S40. Based on the service resilience index and the current number of idle warehouses, after dynamic redundancy compression and safety lower limit protection, the available safety margin that determines the upper limit of the number of serviceable warehouses is obtained.

[0039] S41. Set the basic number of redundant warehouses and the absolute bottom line, and dynamically adjust the redundancy requirements according to the ambient temperature, transforming physical safety constraints into adaptively adjustable scheduling boundary conditions.

[0040] Based on the thermal management design capabilities of the battery swapping cabinet and the thermal runaway propagation characteristics of the batteries used, the basic number of redundant compartments is set. The basic redundancy number of bays is the larger of 2 and 10% of the total number of bays N, rounded up. This value ensures that even in the event of a single-bay thermal runaway, there are still at least two idle bays serving as physical isolation buffers to prevent heat spread and maintain effective heat dissipation. Simultaneously, a minimum of one emergency isolation bay is defined as an absolute safety baseline that cannot be breached; that is, any scheduling strategy must reserve at least one empty bay for emergency isolation. An environmental temperature adaptive mechanism is introduced: when the ambient temperature exceeds a set temperature threshold and enters a high-temperature season, the basic redundancy number of bays automatically increases by 20% to compensate for the additional risks caused by decreased heat dissipation efficiency and increased battery sensitivity at high temperatures. This mechanism explicitly embeds physical safety constraints into the scheduling framework, providing reliable, adjustable, and engineering-realistic boundary conditions for subsequent elastic capacity calculations.

[0041] S42. Based on the service resilience index, the available service ratio of the battery swapping cabinet is dynamically adjusted to achieve risk-adaptive elastic resource scheduling, maximizing service efficiency while ensuring safety.

[0042] The available safety margin is determined dynamically based on the real-time calculated service resilience index, which determines the proportion of battery swapping stations that can provide external services. When the service resilience index is higher than the set resilience threshold, it indicates that the battery swapping station is operating stably and the risks are controllable. In this case, all stations are allowed to participate in user battery swapping services. When the service resilience index is less than or equal to the set resilience threshold, a dynamic redundancy compression mechanism is used, according to the formula... Calculate the available position ratio. This represents the current number of available compartments in the battery swapping station. Based on the number of redundant warehouses, This represents the current service resilience index. This ratio ensures that when security risks increase and there are many idle units, some empty units are proactively frozen as an enhancement of redundancy to prevent the depletion of safety margins due to over-scheduling. Conversely, when the battery swapping cabinet is in a stable state and there are few idle units, resources are released as much as possible to maintain service throughput. The entire process implements an adaptive elastic strategy, maximizing the service efficiency and resource utilization of the cabinets while strictly adhering to physical security boundaries. The above ASM is a preliminary strategy value; the final number of available units needs to be determined after security boundary protection in step S43.

[0043] S43. Implement dual security protection based on the service resilience index, and convert the proportion of available services into the number of executable scheduling warehouses that balance security and efficiency based on the service resilience index.

[0044] After dynamically calculating the available service ratio of the battery swapping stations, a dual safety protection mechanism is immediately implemented: First, if the service resilience index is lower than the minimum resilience threshold, indicating that the system is in an extremely high-risk state, all external battery swapping services are suspended, allowing only battery return or waiting for manual maintenance intervention. Second, if the service resilience index is greater than or equal to the minimum resilience threshold but less than or equal to the set resilience threshold, the available station ratio in step S42 is converted into a specific number of serviceable stations, using the formula... Calculate the actual number of battery compartments available for users to replace batteries. The number of emergency isolation chambers is determined. Based on this, a more refined chamber allocation strategy is generated: fully charged batteries are prioritized for chambers with lower thermoelectric imbalance indices, high-risk areas with high thermoelectric coupling propagation potential are actively avoided, and additional energy sources are prevented from being introduced into locations sensitive to thermal runaway.

[0045] S50: Based on the difference between the available safety margin of the battery swapping cabinet and the service resilience index of neighboring cabinets, the regional collaborative guidance strength driving user diversion is obtained through standard deviation discrimination and exponential decay mapping, thereby achieving load balancing and risk dispersion without a central coordination mechanism.

[0046] S51, the battery swapping cabinet communicates with neighboring cabinets through periodic encrypted broadcasts, combined with smooth processing via sliding windows, to build a lightweight regional resilience state awareness network in a decentralized architecture.

[0047] Each battery swapping station proactively broadcasts its service resilience index value and geographical coordinates every 10 seconds via its built-in communication module. Simultaneously, it continuously monitors similar broadcasts from other battery swapping stations within a 300-meter radius. The received service resilience index data from neighboring stations is smoothed and filtered through a sliding window of length 5 to suppress abnormal fluctuations caused by sensor noise or transient disturbances. All communication messages are embedded with digital signatures and timestamps based on device keys, ensuring data reliability, timeliness, and protection against replay and spoofing attacks. This mechanism does not rely on a central server, autonomously building a low-overhead, highly robust regional state-aware network at the edge, providing a real-time and reliable neighborhood resilience view for subsequent collaborative scheduling.

[0048] S52. By introducing the distance-weighted service resilience index standard deviation, the resilience difference between the battery swapping cabinet and its neighboring cabinets is quantified. When the difference is significant and the battery swapping cabinet has the highest risk, collaborative scheduling is triggered.

[0049] Based on the service resilience index values ​​of the battery swapping station and its neighboring stations, using the formula... Calculate the standard deviation of the regional resilience state, where K is the total number of neighboring units that received valid broadcasts. This represents the current service resilience index value of the battery swapping station. Let h be the service resilience index value of the h-th neighboring counter. This is the reciprocal of the straight-line distance between the battery swapping station and its h-th neighboring station. If this standard deviation exceeds a preset difference threshold and the battery swapping station's service resilience index is the lowest in the region, then a significant resource-risk imbalance is identified, indicating that the battery swapping station is in a state of resource scarcity and high risk, requiring the activation of a collaborative scheduling mechanism. This process transforms spatial resilience differences into clear quantitative criteria, effectively avoiding meaningless battery allocation between stations in similar states.

[0050] S53. After confirming the regional imbalance, based on the difference in service resilience index between the battery swapping cabinet and the neighboring cabinet and its own availability margin, a collaborative guidance strength is generated through nonlinear mapping, and a graded prompt is triggered accordingly to achieve adaptive user diversion driven by risk differences.

[0051] Once the imbalance of regional service resilience index is confirmed, the maximum absolute difference between the service resilience index of the battery swapping station and all its neighboring stations is first calculated. Substituting this difference into the nonlinear decay function generates the cooperative guidance intensity, with a value range of [0,1]. The specific formula is as follows: ,in The control function decay rate is set to 2, and ASM represents the current available safety margin of the battery swapping station. The collaborative guidance strength directly drives the user application's guidance strategy: a strong prompt is displayed when CSF > 0.7, guiding the user to a high-resilience neighboring station; a moderate prompt is displayed when 0.4 < CSF ≤ 0.7, indicating that a better-condition battery swapping station is nearby and the user can consider moving there; and no prompt is displayed when CSF ≤ 0.4. Since the collaborative guidance strength depends on both the available safety margin and the maximum absolute difference in service resilience index, when all stations in the network are in a low service resilience index state (i.e., no healthy neighboring stations can be guided), the maximum absolute difference in service resilience index is very small, causing the collaborative guidance strength to approach 0, thus automatically suppressing invalid migration prompts.

[0052] Example 2 like Figure 2 As shown, Embodiment 2 of this application provides an intelligent operation monitoring system for a battery swapping cabinet, comprising: Thermoelectric Imbalance Module: Based on 24-hour sliding statistical data of voltage and temperature in a single battery compartment within the battery swapping cabinet, a thermoelectric imbalance index is obtained through standardized deviation calculation and exponential smoothing. This index characterizes the thermal and electrical deviations of the batteries in that compartment. Specifically, it is divided into the following sub-modules: Data Acquisition and Statistics Submodule: Real-time acquisition of battery compartment voltage and temperature data, calculation of statistics and sequence sample entropy through exponential decay weighted sliding window, providing a time-sensitive calculation benchmark for thermoelectric imbalance assessment.

[0053] Constructing an index submodule: By standardizing voltage and temperature deviations, introducing dynamic weighting of sample entropy and integrating the exponentially smoothed imbalance deterioration rate, an adaptive thermoelectric imbalance index is constructed to quantify the multidimensional anomaly risks of the battery compartment.

[0054] The triple mechanism submodule transforms the thermoelectric imbalance index value into a reliable sub-healthy ward marker through dynamic threshold, deterioration trend and duration verification, and incorporates it into queue management to support upper-level risk decision-making.

[0055] Thermoelectric Coupling Module: Based on multi-physical information of the warehouse with a thermoelectric imbalance index exceeding the threshold, a multi-dimensional topological node set is constructed. Combined with a minimum spanning tree structure, the thermoelectric coupling propagation potential reflecting the spatial aggregation risk of high-risk warehouses is calculated. Specifically, it is divided into the following sub-modules: Topology Node Submodule: High-risk battery compartments are selected from the sub-healthy compartments, and their spatial coordinates, thermoelectric parameters, internal resistance and cooling status are integrated to construct a multi-dimensional topology node set, providing structured risk source data for thermal propagation calculation.

[0056] Spanning Tree Submodule: Constructs a complete graph based on the physical coordinates of the high-risk battery compartments, and generates a minimum spanning tree using the Prim algorithm to depict the spatial distribution and adjacency relationships of the abnormal compartments in a compact manner.

[0057] Computational Coupling Submodule: Based on the minimum spanning tree structure and multidimensional topological node set, calculate the thermoelectric coupling propagation potential to quantify the potential risk of high-risk inter-warehouse thermal runaway chain reaction.

[0058] Service resilience module: Based on thermoelectric coupling propagation potential and waiting queue length, it integrates time-period adaptive weighting and tail risk penalty to obtain a service resilience index that combines hardware risk and service pressure. Specifically, it is divided into the following sub-modules: The load metrics submodule calculates the instantaneous load density based on the current waiting queue length, and then applies a nonlinear compression function to calculate the normalized load metrics based on the load density, providing a basis for scheduling strategies.

[0059] Dual-weight submodule: Dynamically generates the current load weight and security weight based on historical load data, and performs normalization processing to achieve adaptive security and service collaborative optimization under different loads and time periods.

[0060] The resilience calculation submodule calculates the service resilience index based on thermoelectric coupling propagation potential, normalized load index, waiting queue length, and dual adaptive weights.

[0061] The security margin module, based on the service resilience index and the current number of idle warehouses, uses dynamic redundancy compression and a safety lower limit protection to determine the maximum number of serviceable warehouses. It is specifically divided into the following sub-modules: Redundancy Requirements Submodule: Sets the basic number of redundant warehouses and the absolute bottom line, and dynamically adjusts the redundancy requirements according to the ambient temperature, transforming physical safety constraints into adaptively adjustable scheduling boundary conditions.

[0062] Service Ratio Submodule: Based on the service resilience index, dynamically adjust the available service ratio of the battery swapping cabinet to achieve risk-adaptive elastic resource scheduling and maximize service efficiency while ensuring safety.

[0063] The execution scheduling submodule implements dual security protection based on the service resilience index, and converts the proportion of available services into the number of executable scheduling warehouses that balances security and efficiency based on the service resilience index.

[0064] Collaborative Guidance Module: Based on the available safety margin of the battery swapping station and the difference in service resilience index between neighboring stations, the module uses standard deviation discrimination and exponential decay mapping to obtain the regional collaborative guidance strength driving user diversion, achieving load balancing and risk dispersion without a central coordinator. Specifically, it consists of the following sub-modules: Sensing Network Submodule: The battery swapping cabinet communicates with neighboring cabinets through periodic encrypted broadcasts, combined with smoothing processing via sliding windows, to build a lightweight regional resilience state sensing network in a decentralized architecture.

[0065] Resilience Difference Submodule: The resilience difference between the battery swapping cabinet and its neighboring cabinets is quantified by introducing the standard deviation of the distance-weighted service resilience index. When the difference is significant and the battery swapping cabinet has the highest risk, collaborative scheduling is triggered.

[0066] User diversion submodule: After confirming regional imbalance, based on the difference in service resilience index between the battery swapping cabinet and its neighboring cabinets and its own availability margin, it generates collaborative guidance strength through nonlinear mapping, and triggers graded prompts accordingly to achieve adaptive user diversion driven by risk differences.

[0067] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solution of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for intelligent operation monitoring of a battery swapping cabinet, characterized in that, include: S10. Based on the 24-hour sliding statistical data of voltage and temperature of a single battery compartment in the battery swapping cabinet, a thermoelectric imbalance index is obtained through standardized deviation calculation and exponential smoothing, which is used to characterize the thermal and electrical deviation of the battery in the compartment. S20. Based on the multi-physical information of the warehouse with the thermoelectric imbalance index exceeding the threshold, a multi-dimensional topological node set is constructed. Combined with the minimum spanning tree structure, the thermoelectric coupling propagation potential reflecting the spatial aggregation risk of high-risk warehouses is calculated. S30. Based on thermoelectric coupling propagation potential and waiting queue length, the service resilience index that integrates hardware risk and service pressure is obtained by combining time period adaptive weighting with tail risk penalty. S40. Based on the service resilience index and the current number of idle warehouses, after dynamic redundancy compression and safety lower limit protection, the available safety margin that determines the upper limit of the number of serviceable warehouses is obtained. S50: Based on the difference between the available safety margin of the battery swapping cabinet and the service resilience index of neighboring cabinets, the regional collaborative guidance strength driving user diversion is obtained through standard deviation discrimination and exponential decay mapping, thereby achieving load balancing and risk dispersion without a central coordination mechanism.

2. The intelligent operation monitoring method for a battery swapping cabinet as described in claim 1, characterized in that, Based on 24-hour sliding statistical data of voltage and temperature in a single battery compartment of the battery swapping cabinet, a thermoelectric imbalance index is obtained through standardized deviation calculation and exponential smoothing. This index is used to characterize the thermal and electrical deviation of the battery in that compartment. The process is divided into the following sub-steps: Real-time acquisition of battery compartment voltage and temperature data, and calculation of statistics and sequence sample entropy through exponential decay weighted sliding window, providing a time-sensitive calculation benchmark for thermoelectric imbalance assessment; By standardizing voltage and temperature deviations, introducing dynamic weighting of sample entropy and fusing exponentially smoothed imbalance deterioration rates, an adaptive thermoelectric imbalance index is constructed to quantify the multidimensional anomaly risks of the battery compartment. Through triple verification of dynamic threshold, deterioration trend and duration, the thermoelectric imbalance index value is transformed into a reliable sub-healthy ward marker and incorporated into the cohort management to support upper-level risk decision-making.

3. The intelligent operation monitoring method for a battery swapping cabinet as described in claim 2, characterized in that, Through triple verification of dynamic threshold, deterioration trend, and duration, the thermoelectric imbalance index value is transformed into a reliable sub-healthy status marker and incorporated into cohort management to support upper-level risk decision-making. This process is divided into the following sub-steps: The dynamic threshold mechanism first sets an exponential base threshold, and then superimposes a periodic correction term on the base threshold to simulate the impact of diurnal temperature variation on normal fluctuations. If the current thermoelectric imbalance index is greater than the correction term, the battery compartment is marked as a suspected anomaly, and a trend analysis mechanism is used to verify whether it shows a continuous deterioration trend. Only when the thermoelectric imbalance index is greater than the correction term and shows a continuous deterioration trend will the battery compartment be upgraded to a candidate for sub-health. The duration verification mechanism implements time window cumulative verification for sub-health candidate battery compartments. If the thermoelectric imbalance index of the compartment exceeds the limit for more than 30 minutes, the battery compartment is officially marked as sub-healthy and added to the priority monitoring queue. If the thermoelectric imbalance index value falls below the correction item or does not show a deteriorating trend, the battery compartment returns to a healthy state.

4. The intelligent operation monitoring method for a battery swapping cabinet as described in claim 1, characterized in that, Based on the multi-physical information of the warehouse with a thermoelectric imbalance index exceeding the threshold, a multi-dimensional topological node set is constructed. Combined with the minimum spanning tree structure, the thermoelectric coupling propagation potential reflecting the spatial aggregation risk of high-risk warehouses is calculated. The specific steps are as follows: High-risk battery compartments are selected from sub-healthy compartments, and their spatial coordinates, thermoelectric parameters, internal resistance and cooling status are integrated to construct a multi-dimensional topological node set, providing structured risk source data for thermal spread calculation; A complete graph is constructed based on the physical coordinates of the high-risk battery compartments, and a minimum spanning tree is generated using the Prim algorithm to depict the spatial distribution and adjacency relationships of the abnormal compartments in a compact manner. The potential risk of thermal runaway chain reaction in high-risk warehouses is quantified by calculating the thermoelectric coupling propagation potential based on the minimum spanning tree structure and multidimensional topological node set.

5. The intelligent operation monitoring method for a battery swapping cabinet as described in claim 1, characterized in that, Based on the thermoelectric coupling propagation potential and the waiting queue length, an exponential adjustment of time-period adaptive weights and tail risk penalty are combined to obtain a service resilience index that integrates hardware risk and service pressure. This is specifically divided into the following sub-steps: The instantaneous load density is calculated by the current waiting queue length, and the normalized load index is calculated by applying a nonlinear compression function based on the load density, providing a basis for scheduling strategies. The load weight and security weight for the current moment are dynamically generated based on historical load data and normalized to achieve adaptive security and service co-optimization under different loads and time periods. The service resilience index is calculated based on thermoelectric coupling propagation potential, normalized load index, waiting queue length, and dual adaptive weighting.

6. The intelligent operation monitoring method for a battery swapping cabinet as described in claim 1, characterized in that, Based on the service resilience index and the current number of idle warehouses, after dynamic redundancy compression and safety lower limit protection, the availability safety margin that determines the upper limit of the number of serviceable warehouses is obtained, which is specifically divided into the following sub-steps: Set a basic number of redundant warehouses and an absolute bottom line, and dynamically adjust the redundancy requirements according to the ambient temperature, transforming physical safety constraints into adaptively adjustable scheduling boundary conditions; The available service ratio of the battery swapping cabinet is dynamically adjusted based on the service resilience index to achieve risk-adaptive elastic resource scheduling and maximize service efficiency while ensuring safety. Dual security protection is implemented based on the service resilience index, and the proportion of available services is converted into the number of executable scheduling warehouses that balance security and efficiency based on the service resilience index.

7. The intelligent operation monitoring method for a battery swapping cabinet as described in claim 1, characterized in that, Based on the difference between the available safety margin of the battery swapping station and the service resilience index of neighboring stations, the regional collaborative guidance strength driving user diversion is obtained through standard deviation discrimination and exponential decay mapping. This achieves load balancing and risk dispersion without a central coordinator, specifically through the following sub-steps: The battery swapping cabinet communicates with neighboring cabinets through periodic encrypted broadcasts, combined with smooth processing via sliding windows, to build a lightweight regional resilient state awareness network in a decentralized architecture. By introducing the distance-weighted service resilience index standard deviation, the resilience difference between the battery swapping station and its neighboring stations is quantified. When the difference is significant and the battery swapping station has the highest risk, collaborative scheduling is triggered. After confirming the regional imbalance, based on the difference in service resilience index between the battery swapping cabinet and its neighboring cabinets and its own availability margin, a collaborative guidance strength is generated through nonlinear mapping, and a graded prompt is triggered accordingly to achieve adaptive user diversion driven by risk differences.

8. A smart operation monitoring system for a battery swapping cabinet, characterized in that, include: Thermoelectric Imbalance Module: Based on 24-hour sliding statistical data of voltage and temperature of a single battery compartment in the battery swapping cabinet, a thermoelectric imbalance index is obtained through standardized deviation calculation and exponential smoothing, which is used to characterize the thermal and electrical deviation of the battery in the compartment. Thermoelectric Coupling Module: Based on the multi-physical information of the thermoelectric imbalance index exceeding the threshold, a multi-dimensional topological node set is constructed. Combined with the minimum spanning tree structure, the thermoelectric coupling propagation potential reflecting the spatial aggregation risk of high-risk warehouses is calculated. Service resilience module: Based on thermoelectric coupling propagation potential and waiting queue length, it integrates time-period adaptive weighting of exponential adjustment and tail risk penalty to obtain a service resilience index that combines hardware risk and service pressure. Safety margin module: Based on the service resilience index and the current number of idle warehouses, through dynamic redundancy compression and safety lower limit protection, the available safety margin that determines the upper limit of the number of warehouses that can be served is obtained; Collaborative Guidance Module: Based on the difference between the available safety margin of the battery swapping cabinet and the service resilience index of neighboring cabinets, the regional collaborative guidance intensity driving user diversion is obtained through standard deviation discrimination and exponential decay mapping, thereby achieving load balancing and risk dispersion without a central coordination mechanism.

Images