Capsule cabin equipment health management and predictive maintenance system based on digital twinning

Abstract

The invention discloses a capsule cabin equipment health management and predictive maintenance system based on digital twinning, and relates to the technical field of health management, and the system comprises an LPWS index calculation module which is used for constructing a geometric twinning body of a target capsule cabin team, and calculating the LPWS index of each capsule cabin in the target capsule cabin team according to the geometric twinning body; the maintenance frequency calculation module is used for calculating the planned maintenance frequency of each capsule cabin according to the equivalent confluence width of each capsule cabin, the LPWS index and the management maintenance period of the target capsule cabin team; and the offset index calculation module is used for calculating the static offset index of the capsule cabin according to the actually measured standby current of the capsule cabin and the reference standby current of the capsule cabin. According to the method, the capability of identifying the potential catchment angle and the drainage failure area is improved.

Description

Technical Field

[0001] This invention relates to the field of health management technology, and in particular to a health management and predictive maintenance system for capsule cabin equipment based on digital twins. Background Technology

[0002] In operational environments with densely packed capsule cabins, including large transportation hubs, rail stations, underground passages, and other facilities, capsule cabins are increasingly being deployed as modular devices providing temporary rest, security assistance, and environmental monitoring. Because these cabins are typically arranged along a fixed axis in structural spaces with slight ground slopes, their layout is constrained by the station's spatial shape, construction conditions, and cabin dimensional errors, resulting in slight elevation differences between cabins and creating localized low-lying points, particularly at the ends of the cabin rows where low-lying water collection angles are more likely to form. These low-lying points are highly susceptible to water accumulation under conditions of rainwater leakage, air conditioning condensate, or residual spray liquid, leading to dampness at the cabin's bottom structure, short circuits, material corrosion, or functional failure, severely impacting system stability and passenger experience. Currently, maintenance personnel typically rely on visual inspection and reactive responses, making it difficult to promptly identify the location and risk level of potential hazards. Maintenance scheduling relies on experience-based decisions, resulting in low resource utilization efficiency and significant uncertainty in maintenance outcomes.

[0003] Existing methods often overlook the complex relationship between the low-potential structural state of the hull and the electrical static offset when scheduling hull maintenance tasks. They lack a comprehensive sorting logic that can simultaneously consider water catchment risk, electrical anomalies, and spatial location distribution, resulting in high-risk hulls not being prioritized for identification and maintenance. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies that often overlook the complex correlation between the low-potential structural state of the capsule and the electrical static offset when scheduling capsule maintenance tasks, and to propose a capsule equipment health management and predictive maintenance system based on digital twins.

[0005] To address the problems existing in the prior art, the present invention adopts the following technical solution: A digital twin-based capsule equipment health management and predictive maintenance system includes: The LPWS index calculation module is used to construct a geometric twin of the target capsule fleet and calculate the LPWS index of each capsule in the target capsule fleet based on the geometric twin. The maintenance count calculation module is used to calculate the planned maintenance count for each capsule based on the equivalent bus width, LPWS index, and management and maintenance cycle of the target capsule fleet. The offset index calculation module is used to calculate the static offset index of the capsule based on the measured standby current of the capsule and the reference standby current of the capsule. The priority generation module is used to generate a sorting key vector based on the static offset index, and to determine the final maintenance priority queue of the target capsule team based on the sorting key vector. The maintenance task allocation module is used to perform maintenance on each capsule based on the final maintenance priority queue and the planned number of maintenance sessions.

[0006] Preferably, the geometric twin of the target capsule team includes: Measure the distance along the row of each capsule in the target capsule array in the direction of the capsule row; Obtain the ground slope vector; Obtain the angle between the ground slope vector and the cabin row direction vector of each capsule; Obtain the installation leveling error for each capsule compartment; Obtain the equivalent reduction in each capsule compartment caused by the door widening; The geometric twin of the target capsule team is constructed based on the distance along the row, the ground slope vector, the vector angle, the installation leveling error, and the equivalent reduction.

[0007] Preferably, the LPWS index of each capsule in the target capsule fleet is calculated based on the geometric twin, including: In the coordinate system of the geometric twin, the projection component of the ground slope vector onto the cabin direction vector is calculated based on the ground slope vector and the angle between the vectors. The elevation difference between the two ends of the cabin is obtained by multiplying the projected components and the distance along the deck; The relative low potential of each capsule is obtained by adding the ground elevation difference at both ends of the capsule, the installation leveling error, and the equivalent reduction. The ratio of the relative low potential and the minimum starting head of the gravity drainage trough is calculated to obtain the low potential ratio of each capsule. The difference between 1 and the low potential ratio is calculated to obtain the non-starting ratio. The LPWS index for each capsule is obtained by taking the maximum value between 0 and the non-started ratio.

[0008] Preferably, the planned maintenance frequency for each capsule is calculated based on its equivalent bus width, LPWS index, and the management and maintenance cycle of the target capsule fleet, including: Obtain the equivalent bus width for each capsule. Multiply the equivalent bus width by the LPWS index to obtain the risk weight of each capsule. The risk weights of all capsules are summed to obtain the total risk weights. Divide the risk weight of the capsule by the sum of the risk weights to obtain the risk weight percentage. Obtain the management and maintenance cycle of the target capsule cabin team; Multiply the total number of maintenance slots for the management and maintenance cycle by the risk weight ratio to obtain the maintenance slot allocation value for each capsule. The maintenance quota allocation value is rounded to obtain the planned maintenance number for each capsule.

[0009] Preferably, the static offset index of the capsule is calculated based on the measured standby current of the capsule and the reference standby current of the capsule, including: The actual standby current of the capsule was measured when the capsule was stationary. The difference between the measured standby current and the reference standby current of the capsule is calculated to obtain the standby current deviation; The static offset index of the capsule is obtained by calculating the ratio of the standby current deviation to the reference standby current.

[0010] Preferably, generating the sort key vector based on the static offset index includes: Numerical comparison of relatively low potential and minimum starting head: If the relative low potential is less than the minimum starting head, then it is determined that the capsule corresponding to the relative low potential satisfies the low potential catchment angle condition of the end capsule, and the low potential catchment angle flag of the capsule corresponding to the relative low potential is assigned to 1. If the relative low potential is greater than or equal to the minimum starting head, then the low potential catchment angle flag is set to 0. The position factor is obtained by dividing the distance of each capsule along the row by the total row length of the target capsule fleet. Based on the low-potential catchment angle marker, risk weight, static offset index, and location factor of the capsule, a sorting key vector for the capsule is generated.

[0011] Preferably, the final maintenance priority queue for the target capsule fleet is determined based on the sorting key vector, including: The low-potential catchment angle indicator, risk weight, static offset index, and location factor of the sorting key vector are assigned weights proportionally. The low-potential catchment angle indicator, risk weight, static offset index and position factor are weighted and summed to obtain the comprehensive quantitative value of each capsule. The comprehensive quantification values ​​are sorted in descending order to obtain the final maintenance priority queue.

[0012] Preferably, each capsule is maintained according to the final maintenance priority queue and the planned number of maintenance sessions, including: The maintenance priority of each capsule is sorted according to the final maintenance priority queue; The maintenance quota for each capsule is allocated and resources are scheduled based on the planned maintenance frequency.

[0013] Compared with the prior art, the beneficial effects of the present invention are: 1. In this invention, by constructing a geometric twin of the target capsule cabin team, parameters such as ground slope vector, cabin drainage direction, installation leveling error, and equivalent reduction are introduced to accurately restore the real spatial relationship between the capsule cabins and the ground. This enables the system to dynamically calculate the low-potential state and LPWS index of each cabin, achieving quantitative modeling and visual identification of drainage risks. Compared with the traditional method based on experience judgment and static drawings, the digital twin can globally model the spatial height difference and head change of the cabin drainage in a unified coordinate system, effectively improving the ability to identify potential catchment angles and drainage failure areas, and providing a solid geometric foundation and risk basis for subsequent predictive maintenance.

[0014] 2. In this invention, an index such as LPWS index, equivalent confluence width, and static offset index is combined into a multi-dimensional sorting key vector, realizing a comprehensive sorting mechanism that covers three-dimensional factors of water potential, spatial location, and electrical anomalies. This improves the scientificity and accuracy of maintenance task arrangement, ensures that the maintenance sequence is highly matched with the actual risk distribution, improves the efficiency of maintenance resource utilization, and reduces the probability of missing high-risk compartments. Attached Figure Description

[0015] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 This is a functional block diagram of a capsule cabin equipment health management and predictive maintenance system based on digital twins, provided in an embodiment of the present invention. Detailed Implementation

[0016] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0017] Example: This example provides a capsule cabin equipment health management and predictive maintenance system based on digital twins. See [link to example]. Figure 1 Specifically, including: The LPWS index calculation module is used to construct a geometric twin of the target capsule fleet and calculate the LPWS index of each capsule in the target capsule fleet based on the geometric twin. In an embodiment of the present invention, constructing a geometric twin of the target capsule pod fleet includes: Measure the distance along the row of each capsule in the target capsule array in the direction of the capsule row; Specifically, when measuring the distance along the row of each capsule in the target capsule array, the axis of the row direction is first determined in the geometric coordinate system of the digital twin system. Taking the reference point at the head of the row as the starting point, the coordinates of the center reference point of each capsule are recorded sequentially along the row direction. Then, the projected length from the center point of each capsule to the starting point in the row direction is calculated, and the obtained distance is stored as the distance along the row in the product data management system.

[0018] Obtain the ground slope vector; Specifically, when obtaining the ground slope vector, the elevation of multiple points within the target ground area is measured, and the measured elevation data of each point is fitted and calculated. Alternatively, three-dimensional scanning technology can be used to scan the ground in this area and then calculate back to obtain a vector that has both direction and magnitude. The direction of the vector points to the side where the ground elevation decreases, and the magnitude of the vector represents the change in ground elevation corresponding to each meter of horizontal distance, with the unit being meters per meter.

[0019] Specifically, the distance along the hull refers to the horizontal projection length continuously measured from the starting point of the hull to the end point on a predetermined hull direction axis. It is used to determine the position and relative distance of each capsule on this axis. It is usually obtained by surveying coordinates or laser rulers and is measured in meters. Its value is linearly related to the cumulative elevation difference of the ground in that direction. The ground slope vector is a directional quantity that simultaneously expresses the direction and magnitude of the ground slope. The direction of the vector points to the side where the elevation decreases, and the magnitude of the vector represents the change in elevation corresponding to each meter of horizontal distance, measured in meters per meter. It can be obtained by fitting the elevation of multiple points in the area or by back-calculating from a three-dimensional scan. This vector provides the basis for any subsequent directional projection and head estimation.

[0020] Obtain the angle between the ground slope vector and the cabin row direction vector of each capsule; Specifically, the angle between the ground slope vector and the capsule discharge direction vector describes the orientation difference of the ground descent direction relative to the capsule discharge axis. This angle determines that the effective slope component is equal to the slope magnitude multiplied by the cosine of this angle. When the angle approaches zero, the effective slope along the discharge direction is maximum, and water is more likely to converge at the low-potential end of the capsule discharge. When the angle is close to a right angle, there is almost no cumulative height difference along the discharge direction, making it difficult to form unilateral water collection. When the angle exceeds a right angle, the effective slope along the discharge direction is negative, and the risk shifts from the tail of the discharge to the head of the discharge. Therefore, this angle, together with the distance along the discharge and the ground slope vector, determines the relative potential difference of each capsule bottom frame angle and directly affects whether a low-potential water collection angle appears at the end of the capsule and its strength.

[0021] Obtain the installation leveling error for each capsule compartment; Obtain the equivalent reduction in each capsule compartment caused by the door widening; Specifically, when obtaining the installation leveling error for each capsule, first determine the elevation of the design reference horizontal plane corresponding to the bottom support plane specified in the capsule installation design document. Then, use a high-precision level and laser positioning instrument to measure the actual elevation of multiple reference points on the bottom support plane of each capsule, and record the actual elevation value of each reference point. Subsequently, calculate the difference between the actual elevation value of each reference point and the design reference horizontal plane elevation. Based on the requirement that this difference reflects the slight tilt of the capsule, take the maximum value or arithmetic mean of the differences as the installation leveling error of the capsule. When obtaining the equivalent reduction caused by the widening of the door for each capsule, first... From the initial design drawings of the capsule or the initial design model of the geometric twin of the target capsule team, the design elevation data of the bottom edge of the capsule body before the door widening (including the bottom of the door frame and the surrounding bottom frame) is extracted. Then, the bottom edge of the capsule body after the door widening is scanned by a 3D laser scanning device to obtain the actual elevation data of the corresponding position after the door widening. Next, the difference calculation is performed on the actual elevation data of each corresponding position after the door widening and the design elevation data before the door widening. The difference data of the door frame adjustment area and the bottom frame reduction area are selected. The arithmetic mean or the maximum change value of the difference data in this area is taken as the equivalent reduction of the capsule caused by the door widening.

[0022] Specifically, the installation leveling error refers to the offset of the bottom support plane of the capsule relative to the design reference horizontal plane during installation. This offset reflects the slight tilt of the capsule caused by foundation settlement, uneven pre-tightening of support bolts, or assembly errors during installation. This error will directly affect the consistency between the bottom drainage direction and the drainage direction of the capsule and change the spatial position of the low-potential point. The equivalent reduction refers to the change in the bottom edge elevation of the capsule caused by widening the door or adjusting the door frame structure. When the door is widened, it is usually necessary to lower the bottom frame or adjust the sill to ensure the opening angle and sealing gap of the door. This structural change will cause the local reference height of the capsule to drop relative to the design base plane, thus forming an additional low-potential difference. This low-potential difference can be regarded as the effective height reduction of the capsule in terms of geometry, which has a direct impact on the drainage start-up conditions and water collection angle formation of the end capsule.

[0023] The geometric twin of the target capsule team is constructed based on the distance along the row, the ground slope vector, the vector angle, the installation leveling error, and the equivalent reduction.

[0024] Specifically, when constructing the geometric twin of the target capsule fleet, the following steps are taken: First, the distance along the row axis of each capsule in the target capsule fleet is obtained using a laser ruler or surveying coordinate method, with the starting point of the row as the zero point and the distance to the end point. The ground slope vector is obtained through multi-point elevation fitting within the region or 3D scanning and back-calculation. The vector angle between the ground slope vector and the row direction vector of each capsule is obtained through vector operations. The difference between the actual elevation of the bottom support plane reference point of the capsule and the design reference horizontal plane elevation is measured using a high-precision level instrument, and the statistical value is obtained to determine the installation leveling error. The equivalent reduction is obtained by comparing the difference between the design elevation and the actual elevation of the bottom edge of the capsule before and after the door widening using 3D laser scanning. Next, a virtual coordinate system is established in the digital twin system, with the starting point of the row as the origin, the row direction as the X-axis, the vertical upward direction as the Z-axis, and the horizontal direction perpendicular to the row direction as the Y-axis. Then, the distance along the row of each capsule is assigned to its value in the virtual coordinate system X... The coordinates on the axis are used to calculate the projection component of the ground slope in the X-axis direction based on the ground slope vector and the angle between the vectors, and to determine the tilt trend of the virtual ground. Combined with the installation leveling error, the attitude of the bottom support plane of each capsule in the virtual coordinate system is adjusted to reflect the slight tilt of the capsule. The elevation of the bottom edge of each capsule in the virtual coordinate system on the Z-axis is corrected according to the equivalent reduction to reflect the low potential difference brought about by the structural adjustment. Then, according to the design dimension parameters of each capsule, a corresponding three-dimensional model entity is generated in the virtual coordinate system, so that the spatial position, attitude and elevation of the model entity match the above data one by one. Finally, all the three-dimensional model entities of the capsules are integrated into the virtual coordinate system according to the spatial arrangement of the actual capsules, forming a geometric twin of the target capsule team that can reflect the real geometric relationship between the capsule and the ground, the distribution of relative height difference between the capsules and the potential energy difference, providing a unified geometric basis and simulation reference for subsequent LPWS index calculation, drainage performance analysis and predictive maintenance.

[0025] Specifically, the geometric twin of the target capsule fleet is a three-dimensional geometric counterpart established in the digital twin system based on the spatial arrangement and installation status of the actual capsule fleet, used to reflect the real geometric relationship between the capsules and the ground. This geometric twin generates a model entity with spatial position, attitude, and relative height difference in a virtual coordinate system by collecting data on the distance along the row of each capsule, the ground slope vector, the angle between the capsule and row direction vectors, installation leveling errors, and the equivalent reduction caused by door widening. The system uses this data to calculate the relative height difference distribution and potential energy difference between the capsules, enabling the virtual twin to dynamically reflect the low-potential changes and water catchment characteristics of the row of capsules under different terrain and installation conditions, thus providing a unified geometric basis and simulation reference for LPWS index calculation, drainage performance analysis, and predictive maintenance.

[0026] In general, constructing a geometric twin of the target capsule team aims to fully incorporate the actual geometric relationships of the capsules and their layouts into the system's calculations and operational decisions. The track hub has slight slope differences in the ground level; installation leveling deviations and door widening alter the local elevation of the capsule floor; and gravity drainage channels have a starting threshold head. These factors collectively determine whether a low-potential water collection angle forms at the end and causes water to stagnate under low-volume conditions. Management based solely on ledger parameters cannot quantify the cumulative elevation difference along the drainage path and the projection effect of the ground slope direction, nor can it trace the impact of different construction modifications on the drainage threshold. The geometric twin reconstructs key quantities such as capsule / drainage direction, distance along the drainage path, ground slope vector, angle between two vectors, installation leveling error, and equivalent reduction from door widening using a unified coordinate system. This allows the system to accurately reproduce the real-world elevation difference distribution and potential energy pattern in virtual space, continuously calculate the low-potential risk index of each capsule, locate the end angle prone to water collection, and then integrate risk weights and static offset monitoring results as the basis for sorting and dispatching.

[0027] In an embodiment of the present invention, calculating the LPWS index of each capsule in the target capsule fleet based on the geometric twin includes: In the coordinate system of the geometric twin, the projection component of the ground slope vector onto the cabin direction vector is calculated based on the ground slope vector and the angle between the vectors. Specifically, in the coordinate system of the geometric twin, when calculating the projection component of the ground slope vector onto the cabin direction vector based on the ground slope vector and the angle between the vectors, the system first calls the established standardized expressions for the ground slope vector and the cabin direction vector in the coordinate system. The two vectors are then normalized to eliminate scale effects. Next, the cosine value of the angle between the two vectors is calculated through inner product operations. The system then multiplies this cosine value with the modulus of the ground slope vector to obtain the effective component of the ground slope along the cabin direction. This calculation process is automatically executed by the geometric analysis module of the digital twin system. The modulus of the ground slope vector represents the rate of change of elevation per unit length, and the cabin direction vector represents the axial direction of the cabin formation. This projection calculation quantitatively obtains the effective slope along the cabin direction, providing basic input parameters for further calculations of the elevation difference between the first and last cabins and the relative low-lying area of ​​the last cabin, and ensuring that all calculation results are traceable and repeatable within a unified coordinate system.

[0028] The elevation difference between the two ends of the cabin is obtained by multiplying the projected components and the distance along the deck; Specifically, the projection component is the effective projection of the ground slope vector onto the cabin direction vector in the angle relationship between the ground slope vector and the cabin direction vector. Its value is the magnitude of the ground slope vector multiplied by the cosine of the angle between the vectors, and it is used to quantify the degree of influence of the ground slope on the potential energy distribution in the cabin direction. The ground elevation difference at both ends of the cabin is the result of multiplying the projection component by the distance along the cabin, representing the difference in ground elevation between the start and end of the cabin in the direction along the cabin. It is the core parameter for determining the formation and strength of the low-potential water collection angle of the capsule cabin, and directly affects the water collection and drainage characteristics between the cabins.

[0029] The relative low potential of each capsule is obtained by adding the ground elevation difference at both ends of the capsule, the installation leveling error, and the equivalent reduction. The ratio of the relative low potential and the minimum starting head of the gravity drainage trough is calculated to obtain the low potential ratio of each capsule. Specifically, the relative low potential value refers to the comprehensive height difference formed by the sum of the height difference caused by the ground slope, the installation leveling error, and the equivalent structural reduction in the direction of each capsule in the direction of the capsule row. This value is used to reflect the potential energy state of the bottom of the capsule relative to the average reference plane of the capsule row. The larger the value, the lower the bottom of the capsule is, and the easier it is for water to accumulate or form a water stagnation angle.

[0030] Specifically, gravity drainage channels are unpressurized drainage channels located at the bottom of capsule cabins or below the ground level of rail transit hubs. They rely on the gravitational potential energy created by the difference in elevation to drive the water to flow out by gravity, requiring no additional power equipment. These drainage channels typically consist of a channel body, an inlet, an air seal section, and an outlet. The longitudinal slope of the channel body determines the flow direction and velocity of the water, while the relative height difference between the inlet elevation and the lowest point of the channel body constitutes the minimum starting head of the drainage system. When condensation, leakage, or spray residue appears in the environment, the water will first collect at the lowest point of the ground. Once the water level reaches the starting head, the water will flow along the drainage channel to the outlet under gravity and be discharged, thus maintaining the dryness and safety of the cabin bottom.

[0031] Specifically, the minimum starting head of the gravity drainage channel refers to the lowest water level difference required for the drainage channel to begin gravity flow. This head value depends on factors such as the drainage channel's geometry, inlet height, flow resistance coefficient, and outlet back pressure. When the water level does not reach this height, continuous drainage cannot be formed in the channel due to insufficient air seal or static pressure, and the water will remain in the low-potential zone. The low-potential ratio is a dimensionless parameter calculated by comparing the relative low-potential amount calculated for the hull with the minimum starting head. It is used to describe the relative relationship between the hull low-potential depth and the drainage starting conditions. When the low-potential ratio is less than one, it indicates that the hull low-potential difference is insufficient to overcome the starting head of the drainage channel, and the water is not easily discharged. When the low-potential ratio is close to or exceeds one, it indicates that the hull has the conditions to start drainage and can achieve gravity flow discharge. Therefore, this ratio is a key criterion for evaluating whether the low-potential catchment angle of the end hull has formed and its severity. It is also a core quantitative indicator for judging the risk of drainage failure in the capsule equipment health management and predictive maintenance system.

[0032] The difference between 1 and the low potential ratio is calculated to obtain the non-starting ratio. The LPWS index for each capsule is obtained by taking the maximum value between 0 and the non-started ratio.

[0033] Specifically, the non-start ratio refers to the numerical result obtained by calculating the difference between the unit value and the low potential ratio of the capsule. It is used to characterize the degree of water stagnation in the capsule when the gravity drainage trough has not been activated. The larger the value, the greater the difference between the bottom water head and the drainage start water head, the more difficult it is to activate the drainage system, and the more likely the capsule is to accumulate water or become damp.

[0034] Specifically, the LPWS index is a comprehensive quantitative indicator used to characterize the risk of low-potential water sinking at the end of the capsule drainage system. LPWS is an abbreviation for Low-Potential-Water-Sink, representing the low-potential water sink angle. The LPWS index is calculated based on a geometric twin constructed from geometric parameters such as the ground slope vector, drainage direction, installation leveling error, and the equivalent reduction caused by door widening. By comparing the relative low potential at the bottom of the capsule with the minimum head required to start the gravity drainage channel, it measures whether the bottom of the capsule is in a stagnant or undrained state. The LPWS index ranges from zero to one. A higher value indicates a lower geometric potential energy at the bottom of the capsule, making it more prone to forming localized water accumulation areas under low water volume conditions. When the index is close to zero, it indicates that the bottom of the capsule already has gravity drainage capabilities and there is no risk of water sinking. This index comprehensively reflects the spatial impact of capsule installation and terrain conditions on drainage performance and is an important basis for identifying drainage hazards, determining maintenance priorities, and generating specific work orders in the capsule equipment health management and predictive maintenance system.

[0035] It should be noted that a low-potential water collection angle refers to a localized lowest point in the bottom area of ​​a capsule or similar modular equipment, formed by the combined effects of the ground slope, installation leveling errors, and the geometry of the capsule structure. Under conditions of small water volumes or leakage, water will preferentially collect and temporarily stagnate at this location, exhibiting a corner characteristic of concentrated water flow and delayed drainage. This area is typically located at the end of the capsule or near the lower edge of the door frame. When the angle between the ground slope and the drainage direction is small, the superposition of installation errors and the lowering effect of the door frame along the drainage slope will create a micro-depression area relatively lower than the drainage channel inlet. Water cannot drain before reaching the starting head of the drainage channel, resulting in water stagnation. The presence of a low-potential water collection angle not only leads to increased local humidity and material aging but also affects the insulation of electrical components and standby power consumption.

[0036] Specifically, the calculation of the LPWS index follows the basic laws of hydrostatics and gravitational potential energy distribution. The projection component of the ground slope vector in the direction of the hopper reflects the elevation gradient corresponding to a unit horizontal length. By multiplying it by the distance along the hopper, the elevation difference between the two ends of the hopper can be obtained. This elevation difference represents the change in gravitational potential energy in the area where the hopper is located. Adding this elevation difference to the installation leveling error and the equivalent structural reduction is equivalent to introducing construction deviations and local geometric settlement into the overall potential energy model, thereby obtaining the relative low potential of the hopper. Then, the ratio of the relative low potential to the minimum starting head of the drainage channel is calculated, which converts the geometric potential energy difference into the form of hydraulic potential energy difference, used to measure whether the potential difference at the bottom of the hopper can drive gravity drainage. When the ratio is less than one, it means that the potential energy at the bottom of the hopper is not enough to overcome the static pressure difference required to start the drainage channel, and the water will stagnate and form a catchment area; when the ratio is close to or greater than one, the potential energy at the bottom of the hopper can stimulate gravity flow, and the risk of water stagnation is eliminated. By calculating the difference between the ratio of 1 and low potential and the maximum value of the ratio of zero and non-starting, hulls below the discharge threshold can be mapped to a state with a value greater than zero, while hulls that have reached the discharge condition can be mapped to zero, thus constructing a continuous and quantifiable low-potential catchment risk index. This index utilizes the principle of energy conservation and the condition of static pressure balance to achieve a physical mapping from geometric spatial difference to discharge risk intensity, and is an effective parameter for identifying the low-potential catchment angle characteristics of terminal hulls.

[0037] The maintenance count calculation module is used to calculate the planned maintenance count for each capsule based on the equivalent bus width, LPWS index, and management and maintenance cycle of the target capsule fleet. In embodiments of the present invention, the planned maintenance frequency for each capsule is calculated based on the equivalent bus width, LPWS index, and management and maintenance cycle of the target capsule fleet, including: Obtain the equivalent bus width for each capsule. Specifically, when obtaining the equivalent runoff width of each capsule, the system first determines the angle between the capsule layout direction and the ground slope direction in the coordinate system of the geometric twin, and then calculates the surface runoff area corresponding to each capsule based on the directional distribution of the ground slope vector. The system identifies the position of the capsule's center point on the capsule layout axis, using the midpoint between adjacent capsules as the runoff boundary line to determine the runoff projection range of a single capsule in the capsule layout direction. Simultaneously, by combining the lateral component of the ground slope vector, the system calculates the lateral diffusion influence width of surface runoff, ensuring that the runoff range takes into account factors such as inter-capsule gaps, slope deflection, and structural obstruction. Based on this, the digital twin system performs an area equivalence transformation on the runoff range, dividing the capsule's runoff area along the capsule layout direction by the capsule length to obtain the equivalent runoff width with length dimensions.

[0038] Multiply the equivalent bus width by the LPWS index to obtain the risk weight of each capsule. The risk weights of all capsules are summed to obtain the total risk weights. Specifically, the equivalent catchment width (ECW) refers to the effective lateral width of each capsule in the drain direction that it can receive and collect surface runoff or seepage water. This width is determined by the bottom shape of the capsule, the gap between adjacent capsules, and the slope of the ground, reflecting the capsule's contribution to the water flow distribution. Risk weight is a quantitative indicator obtained by multiplying the equivalent catchment width by the LPWS index. It represents the relative importance of a single capsule in the drainage risk system. When the capsule's catchment width is large and the LPWS index is high, its risk weight value increases accordingly, indicating that the capsule is more prone to waterlogging or dampness under low water volume conditions. The total risk weight is the overall result obtained by summing the risk weights of all capsules. It characterizes the cumulative drainage risk level of the entire drain or system. A larger value indicates a more concentrated distribution of low-potential areas and more potential drainage bottlenecks in the system, serving as an important basis for maintenance resource allocation and priority assessment.

[0039] Divide the risk weight of the capsule by the sum of the risk weights to obtain the risk weight percentage. Obtain the management and maintenance cycle of the target capsule cabin team; Specifically, the risk weight ratio refers to the ratio of the risk weight value of a single capsule to the sum of the risk weights of the entire capsule row or system, used to characterize the relative influence of that capsule in the overall drainage risk structure. This ratio reflects the proportion of the capsule's contribution to the overall waterlogging risk under low-potential catchment conditions. When a capsule has a large catchment width and a high LPWS index, its risk weight ratio increases accordingly, meaning that the capsule's impact on the overall safe operation of the system is more significant. The management and maintenance cycle refers to the time interval between periodic inspections and maintenance required to maintain the target capsule fleet in a healthy state under given operating conditions. This cycle is determined by comprehensively considering the aging rate of the capsule structure, changes in environmental humidity, the operating frequency of the drainage system, and the distribution of risk weights. It is used to guide the system to dynamically adjust the maintenance frequency of each capsule according to the risk weight ratio, achieving risk-driven management and scheduling, allocating resources to capsules with higher risk ratios, thereby improving overall maintenance efficiency and reliability.

[0040] Multiply the total number of maintenance slots for the management and maintenance cycle by the risk weight ratio to obtain the maintenance slot allocation value for each capsule. The maintenance quota allocation value is rounded to obtain the planned maintenance number for each capsule.

[0041] Specifically, the total maintenance quota for a maintenance cycle refers to the total number of maintenance sessions or maintenance resources that the system can allocate within a complete maintenance cycle. This number is determined comprehensively based on factors such as operational capabilities, staffing, time windows, and cost budgets, and is used to constrain the total scale of the maintenance plan. The capsule maintenance quota allocation value refers to the maintenance amount allocated to each capsule proportionally from the total maintenance quota based on the risk weight of each capsule. This value reflects the resource priority of each capsule in the risk-driven maintenance strategy; a higher value indicates a higher maintenance frequency for that capsule. The planned maintenance sessions for a capsule are integer results obtained by rounding the allocated maintenance quota value. These sessions determine the actual number of maintenance executions for each capsule within the current maintenance cycle. This number directly guides the system's maintenance scheduling and task assignment, ensuring that limited maintenance resources are optimally allocated according to the risk-priority principle.

[0042] Specifically, the calculation of planned maintenance times follows the fundamental principles of system risk distribution and resource conservation. The total number of maintenance slots within a cycle can be considered a finite energy or resource pool. The risk weight of each capsule is equivalent to the spatial distribution density of that resource, reflecting the risk contribution intensity of each capsule in the overall system. Multiplying the total number of maintenance slots by the risk weight ratio achieves a proportional allocation from global maintenance resources to individual capsule maintenance needs, ensuring that maintenance resources flow and are distributed according to the principle of risk energy distribution. Subsequently, the allocated maintenance slot values ​​are rounded, effectively discretizing the continuous quantity into executable maintenance task counts, guaranteeing that each capsule receives work opportunities matching its risk level during the actual maintenance cycle. This calculation process embodies the ideas of energy conservation and risk equilibrium, enabling the system to achieve optimal resource allocation through risk-driven proportional scheduling under the condition of unchanged overall maintenance capacity, thereby obtaining planned maintenance times that conform to the actual risk distribution.

[0043] The offset index calculation module is used to calculate the static offset index of the capsule based on the measured standby current of the capsule and the reference standby current of the capsule. In an embodiment of the present invention, the static offset index of the capsule is calculated based on the measured standby current of the capsule and the reference standby current of the capsule, including: The actual standby current of the capsule was measured when the capsule was stationary. The difference between the measured standby current and the reference standby current of the capsule is calculated to obtain the standby current deviation; Specifically, the measured standby current refers to the actual current consumption of the capsule's electrical system when it is in a static or non-operating state, maintaining basic monitoring, lighting, and environmental control functions. This current reflects the overall health and energy consumption level of the capsule's electrical circuits. The reference standby current is the stable standby current standard value obtained from factory testing or long-term operation of the system, representing the typical static energy consumption of the capsule under normal, fault-free conditions. The standby current deviation is the difference between the measured standby current and the reference standby current, used to describe the degree of deviation of the current electrical system's energy consumption from the normal state. Its positive or negative changes can indicate the changing trends of electrical component aging, insulation moisture, or power line losses.

[0044] The static offset index of the capsule is obtained by calculating the ratio of the standby current deviation to the reference standby current.

[0045] Specifically, the static offset index is a dimensionless parameter calculated by comparing the standby current deviation with the reference standby current. It is used to quantify the relative deviation of the cabin's electrical state. A larger value indicates that the internal electrical system of the cabin is more likely to be affected by moisture or have abnormal power consumption. This index can establish a correlation between electrical performance degradation and the environmental water catchment effect, and is an important basis for assessing moisture risk and determining maintenance priorities in capsule cabin health management.

[0046] Specifically, the process of measuring the standby current of the capsule under static conditions involves using a current sensor to collect the static current of the main electrical circuit in real time when the capsule is stopped and all dynamic loads are disconnected, and recording the average current value during this period. This value reflects the baseline energy consumption level when the capsule is in a stable state. Subsequently, the difference between the measured standby current and the reference standby current formed by factory calibration or long-term monitoring is calculated to obtain the standby current deviation, which represents the degree of deviation of the current system from the standard energy consumption. Finally, by calculating the ratio of the deviation value to the reference standby current, the static offset index of the capsule is obtained, which is used to measure the degree of degradation of the capsule's electrical performance and the impact of environmental humidity. The entire process logically follows the principles of energy conservation and deviation analysis: first, the current energy consumption baseline of the system is determined through measurement; then, abnormal deviations are identified by comparing with the benchmark data; finally, a quantitative characterization of the electrical health status inside the capsule is achieved.

[0047] Specifically, the reason for calculating the ratio of standby current deviation to reference standby current is that this ratio can transform the absolute difference in electrical system energy consumption into a relative rate of change, thereby eliminating the dimensional influence caused by differences in design power, line length, or sensor sensitivity between different capsules. By normalizing with the reference standby current as the denominator, the resulting ratio directly reflects the degree of deviation of current electrical energy consumption relative to the normal state. When the ratio is zero, it indicates that the capsule is in a healthy steady state, while an increase in the ratio indicates additional energy consumption or leakage current in the system, manifested as static energy consumption deviation. This calculation method, based on the proportional relationship between Ohm's law and the power loss formula, can accurately characterize the impact of internal electrical performance degradation, insulation aging, or humidity penetration on current consumption under conditions unaffected by external load fluctuations. Therefore, this ratio is defined as the static deviation index of the capsule, used to achieve quantitative diagnosis and long-term trend monitoring of energy consumption anomalies.

[0048] The priority generation module is used to generate a sorting key vector based on the static offset index, and to determine the final maintenance priority queue of the target capsule team based on the sorting key vector. In an embodiment of the present invention, generating a sort key vector based on a static offset index includes: Numerical comparison of relatively low potential and minimum starting head: If the relative low potential is less than the minimum starting head, then it is determined that the capsule corresponding to the relative low potential satisfies the low potential catchment angle condition of the end capsule, and the low potential catchment angle flag of the capsule corresponding to the relative low potential is assigned to 1. Specifically, the low-potential catchment angle indicator is a state variable used to identify whether the capsule meets the conditions for the formation of a low-potential catchment angle in the terminal compartment. This indicator is determined by comparing the relative low potential of the compartment with the minimum head required to start the drainage channel. When the relative low potential is less than the minimum head required to start, it indicates that the gravitational potential energy at the bottom of the compartment is insufficient to overcome the hydrostatic resistance of the drainage channel, and water will stagnate in this area, forming a localized catchment angle, thus constituting a low-potential catchment angle. The system assigns a value of one to the low-potential catchment angle indicator for the corresponding compartment. This indicator value reflects the critical state of the compartment in terms of geometric potential energy distribution. A value of zero indicates that the bottom of the compartment has normal drainage conditions, while a value of one indicates that the bottom of the compartment is in a state of potential water stagnation risk. This parameter is used in the system to identify the distribution location of low-potential catchment risks in the terminal compartment and serves as a logical input for calculating the LPWS index and maintenance priority ranking, realizing an automated mapping from geometric spatial characteristics to drainage hazard identification.

[0049] Specifically, when the relative low potential is less than the minimum starting head, it indicates that the gravitational potential energy generated by the geometrical elevation difference at the bottom of the compartment is insufficient to overcome the minimum static pressure difference required to initiate water flow in the drainage channel. In other words, the water potential at the bottom of the compartment has not yet reached the critical condition for triggering gravity flow. In this situation, continuous liquid flow cannot be formed in the drainage channel, and water will stagnate and accumulate in local low-lying areas at the bottom of the compartment. With the superposition of the ground slope and structural reduction, this area will become a stable depression for collecting external leakage water or condensate, thus forming a typical low-potential water collection angle. In this state, the water surface of the compartment is in static equilibrium due to the limitation of gravitational potential energy, and cannot be discharged or evaporated, which can easily lead to dampness, mold, or electrical short circuit hazards. Therefore, by judging whether the relative low potential is less than the minimum starting head, it is possible to directly determine whether the compartment is in a low-potential water collection state based on the principle of energy balance. This judgment condition becomes an effective criterion for identifying potential water stagnation risks in the terminal compartment.

[0050] If the relative low potential is greater than or equal to the minimum starting head, then the low potential catchment angle flag is set to 0. Specifically, when the relative low potential is greater than or equal to the minimum starting head, it indicates that the gravitational potential energy corresponding to the geometric height difference at the bottom of the hull has reached or exceeded the minimum static pressure required for gravity flow in the drainage channel. At this point, the water at the bottom of the hull can overcome the flow channel resistance and drain smoothly under the action of gravity. The drainage channel enters a stable flow guiding state, and the bottom of the hull no longer has the conditions to form water accumulation points. Since drainage can proceed naturally, the water potential in the local area will quickly return to equilibrium, and there will be no continuous stagnant water or water accumulation on the ground surface. Therefore, this hull does not belong to the low potential catchment area. In order to logically distinguish the states of different hulls in the system, the system assigns the low potential catchment area flag of such hulls with normal drainage capacity to a value of zero, so that they are identified as a no-catchment-risk state in subsequent risk calculations and maintenance priority assessments, thereby ensuring the accuracy and pertinence of drainage health monitoring and predictive maintenance strategies.

[0051] The position factor is obtained by dividing the distance of each capsule along the row by the total row length of the target capsule fleet. Based on the low-potential catchment angle marker, risk weight, static offset index, and location factor of the capsule, a sorting key vector for the capsule is generated.

[0052] Specifically, the position factor is a normalized geometric parameter used to characterize the spatial distribution of each capsule relative to the overall queue in the direction of the capsule row. It is obtained by dividing the distance of the capsule along the row by the total length of the row, reflecting the relative positional characteristics of the capsule in the overall arrangement. The sorting key vector is a set of multi-dimensional indicators that quantifies the comprehensive state of each capsule. This vector comprehensively considers factors such as low-potential catchment angle indicators, risk weights, static offset indicators, and position factors, forming a composite sorting benchmark that can comprehensively reflect the operational health and potential risk level of the capsules.

[0053] Specifically, the four basic quantities of each capsule are aggregated sequentially and standardized under the same dimensions: First, the low-potential catchment angle marker is read as the primary key, with a value of one or zero to distinguish whether there is a low-potential catchment risk; second, the risk weight of the capsule is read and normalized to a secondary key between zero and one using the maximum risk weight within the same group as the benchmark, to reflect its relative intensity in the overall risk; third, the static offset index is read and normalized to a tertiary key between zero and one using the maximum static offset index within the same group as the benchmark, to express the relative magnitude of electrical static anomalies; fourth, the calculated position factor is used as the quaternary key to describe the relative position of the capsule on the layout axis. Then, the capsule's sorting key vector is assembled in the order of the primary key to the quaternary key.

[0054] In an embodiment of the present invention, determining the final maintenance priority queue of the target capsule fleet based on the sorting key vector includes: The low-potential catchment angle indicator, risk weight, static offset index, and location factor of the sorting key vector are assigned weights proportionally. The low-potential catchment angle indicator, risk weight, static offset index and position factor are weighted and summed to obtain the comprehensive quantitative value of each capsule. The comprehensive quantification values ​​are sorted in descending order to obtain the final maintenance priority queue.

[0055] Specifically, when allocating weights to the low-potential catchment angle marker, risk weight, static offset index, and location factor of the sorting key vector, fault records, maintenance reports, and catchment risk assessment data of the target capsule teams over the past three years are first collected. The correlation coefficients between each element and the probability of capsule failure are statistically analyzed. Combined with scores from at least five industry experts on the impact of each element (out of 10), weights are calculated using the analytic hierarchy process (AHP). The low-potential catchment angle marker is assigned a 40% weight due to its direct association with core waterlogging corrosion risk; the risk weight is assigned a 30% weight due to its contribution to the overall system risk; the static offset index is assigned a 20% weight due to its association with circuit anomaly risk; and the location factor is assigned a 10% weight due to its impact on maintenance convenience. The sum of these four weights is 1. When weighted summing these four elements to obtain the comprehensive quantitative value for each capsule, each element is first standardized. The low-potential catchment angle marker is represented by 0 for no low-potential catchment angle and 1 for the presence of a low-potential catchment angle. The risk weight is taken as the original quantitative value from 0 to 1. The static offset index is calculated based on the measured deviation value and the allowable deviation. The ratio of the difference threshold is calculated (if it exceeds the threshold, it is counted as 1). The position factor is calculated as the ratio of the distance along the cabin row to the total length of the cabin row (range 0 to 1). Then, the standardized value of the low-potential catchment angle of each capsule is multiplied by 40%, the standardized value of the risk weight is multiplied by 30%, the standardized value of the static offset index is multiplied by 20%, and the standardized value of the position factor is multiplied by 10%. The four products are summed, and the sum is the comprehensive quantitative value of the capsule. When sorting the comprehensive quantitative values ​​in descending order to obtain the final maintenance priority queue, a quicksort algorithm is used. The comprehensive quantitative value of each capsule is used as the sorting basis. The comprehensive quantitative value of any capsule in the queue is selected as the benchmark value. All capsules are divided into two groups: one with a comprehensive quantitative value greater than the benchmark value and the other with a comprehensive quantitative value less than the benchmark value. The same sorting operation is recursively performed on the two groups until all capsules are sorted from high to low according to their comprehensive quantitative values. After sorting, the capsules are listed in order. The capsules at the top of the list have higher comprehensive quantitative values ​​and more urgent maintenance needs, thus forming the final maintenance priority queue of the target capsule team.

[0056] The maintenance task allocation module is used to perform maintenance on each capsule based on the final maintenance priority queue and the planned number of maintenance sessions.

[0057] In an embodiment of the present invention, maintenance is performed on each capsule based on the final maintenance priority queue and the planned number of maintenance cycles, including: The maintenance priority of each capsule is sorted according to the final maintenance priority queue; The maintenance quota for each capsule is allocated and resources are scheduled based on the planned maintenance frequency.

[0058] Specifically, when prioritizing each capsule based on the final maintenance priority queue, the order and comprehensive quantitative value of each capsule in the final maintenance priority queue are first extracted. All capsules are then classified into four priorities from front to back: Special, High, Medium, and Basic. The first 15% of the queue are Special, 15% to 40% are High, 40% to 80% are Medium, and the last 20% are Basic. Each priority is assigned a unique identifier: Special is P1, High is P2, Medium is P3, and Basic is P4. This identifier directly relates to the urgency of maintenance operations. When allocating maintenance quotas and scheduling resources based on the planned maintenance frequency for each capsule, the planned maintenance frequency for each capsule is directly used as its maintenance quota, without secondary allocation. Subsequently, resource scheduling rules are formulated based on the priority identifier and quota frequency. For Special priority capsules with a quota frequency of ≥5 times, each maintenance session is staffed with two senior engineers with over 10 years of experience, equipped with high-precision laser leak detectors, intelligent corrosion monitoring terminals, etc. For high-precision equipment, the maintenance time for a single session is set at 1.5 times the standard time, with at least one maintenance window scheduled per week. For high-priority equipment with a quota of 3-4 sessions, each session is staffed by one senior engineer and one intermediate technician, using conventional drainage system detectors and humidity recorders, with a single session duration of 1.2 times the standard time, and a maintenance window scheduled every two weeks. For medium-priority equipment with a quota of 2 sessions, two intermediate technicians operate basic testing tools, with a single session duration of 1.0 times the standard time, and a maintenance window scheduled monthly. For low-priority equipment with a quota of 1 session, one junior technician completes the task with general tools, with a single session duration of 0.8 times the standard time, and a maintenance window scheduled quarterly. A quota consumption log is also established; each completed maintenance session deducts the corresponding capsule's quota until the quota is exhausted. This ensures that the type of maintenance resources, personnel qualifications, time windows, and the priority ranking of each capsule, as well as the determined planned maintenance sessions, are precisely matched, achieving the scheduling goal of orderly progress according to priority and completion of maintenance according to quota.

[0059] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A capsule cabin equipment health management and predictive maintenance system based on digital twins, characterized in that, include: The LPWS index calculation module is used to construct a geometric twin of the target capsule fleet and calculate the LPWS index of each capsule in the target capsule fleet based on the geometric twin. The maintenance count calculation module is used to calculate the planned maintenance count for each capsule based on the equivalent bus width, LPWS index, and management and maintenance cycle of the target capsule fleet. The offset index calculation module is used to calculate the static offset index of the capsule based on the measured standby current of the capsule and the reference standby current of the capsule. The priority generation module is used to generate a sorting key vector based on the static offset index, and to determine the final maintenance priority queue of the target capsule team based on the sorting key vector. The maintenance task allocation module is used to perform maintenance on each capsule based on the final maintenance priority queue and the planned number of maintenance sessions.

2. The capsule cabin equipment health management and predictive maintenance system based on digital twins according to claim 1, characterized in that, Constructing a geometric twin of the target capsule team includes: Measure the distance along the row of each capsule in the target capsule array in the direction of the capsule row; Obtain the ground slope vector; Obtain the angle between the ground slope vector and the cabin row direction vector of each capsule; Obtain the installation leveling error for each capsule compartment; Obtain the equivalent reduction in each capsule compartment caused by the door widening; The geometric twin of the target capsule team is constructed based on the distance along the row, the ground slope vector, the vector angle, the installation leveling error, and the equivalent reduction.

3. The capsule cabin equipment health management and predictive maintenance system based on digital twins according to claim 2, characterized in that, The LPWS index of each capsule in the target capsule fleet is calculated based on the geometric twin, including: In the coordinate system of the geometric twin, the projection component of the ground slope vector onto the cabin direction vector is calculated based on the ground slope vector and the angle between the vectors. The elevation difference between the two ends of the cabin is obtained by multiplying the projected components and the distance along the deck; The relative low potential of each capsule is obtained by adding the ground elevation difference at both ends of the capsule, the installation leveling error, and the equivalent reduction. The ratio of the relative low potential and the minimum starting head of the gravity drainage trough is calculated to obtain the low potential ratio of each capsule. The difference between 1 and the low potential ratio is calculated to obtain the non-starting ratio. The LPWS index for each capsule is obtained by taking the maximum value between 0 and the non-started ratio.

4. The capsule cabin equipment health management and predictive maintenance system based on digital twins according to claim 1, characterized in that, Based on the equivalent bus width, LPWS index, and management and maintenance cycle of the target capsule fleet for each capsule, calculate the planned maintenance frequency for each capsule, including: Obtain the equivalent bus width for each capsule. Multiply the equivalent bus width by the LPWS index to obtain the risk weight of each capsule. The risk weights of all capsules are summed to obtain the total risk weights. Divide the risk weight of the capsule by the sum of the risk weights to obtain the risk weight percentage. Obtain the management and maintenance cycle of the target capsule cabin team; Multiply the total number of maintenance slots for the management and maintenance cycle by the risk weight ratio to obtain the maintenance slot allocation value for each capsule. The maintenance quota allocation value is rounded to obtain the planned maintenance number for each capsule.

5. The capsule cabin equipment health management and predictive maintenance system based on digital twins according to claim 1, characterized in that, Based on the measured standby current of the capsule and the reference standby current of the capsule, the static offset index of the capsule is calculated, including: The actual standby current of the capsule was measured when the capsule was stationary. The difference between the measured standby current and the reference standby current of the capsule is calculated to obtain the standby current deviation; The static offset index of the capsule is obtained by calculating the ratio of the standby current deviation to the reference standby current.

6. The capsule cabin equipment health management and predictive maintenance system based on digital twins according to claim 3, characterized in that, Generate sorted key vectors based on static offset metrics, including: Numerical comparison of relatively low potential and minimum starting head: If the relative low potential is less than the minimum starting head, then it is determined that the capsule corresponding to the relative low potential satisfies the low potential catchment angle condition of the end capsule, and the low potential catchment angle flag of the capsule corresponding to the relative low potential is assigned to 1. If the relative low potential is greater than or equal to the minimum starting head, then the low potential catchment angle flag is set to 0. The position factor is obtained by dividing the distance of each capsule along the row by the total row length of the target capsule fleet. Based on the low-potential catchment angle marker, risk weight, static offset index, and location factor of the capsule, a sorting key vector for the capsule is generated.

7. The capsule cabin equipment health management and predictive maintenance system based on digital twins according to claim 6, characterized in that, The final maintenance priority queue for the target capsule fleet is determined based on the sort key vector, including: The low-potential catchment angle indicator, risk weight, static offset index, and location factor of the sorting key vector are assigned weights proportionally. The low-potential catchment angle indicator, risk weight, static offset index and location factor are weighted and summed to obtain the comprehensive quantitative value of each capsule. The comprehensive quantification values ​​are sorted in descending order to obtain the final maintenance priority queue.

8. The capsule cabin equipment health management and predictive maintenance system based on digital twins according to claim 1, characterized in that, Maintenance is performed on each capsule based on the final maintenance priority queue and the planned number of maintenance sessions, including: The maintenance priority of each capsule is sorted according to the final maintenance priority queue; The maintenance quota for each capsule is allocated and resources are scheduled based on the planned maintenance frequency.

Images