Pipeline connection status prediction system based on digital twin

Abstract

This invention relates to the field of new energy vehicle battery thermal management and digital twin condition monitoring technology, and particularly to a pipeline connection condition prediction system based on digital twins. The system includes a pulse excitation module that issues micro-torque pulse commands to drive servo electric drive components to output alternating micro-torque and extract the pulse response sequence; a parameter acquisition module that retrieves parameters such as axial displacement, rotational deflection, ambient temperature, and vibration frequency; a timing synchronization module that aligns the heterogeneous parameters with the pulse response sequence to generate a timing synchronization parameter set; a signal decoupling module that performs noise reduction and decoupling on the set, extracting the micro-torque elastic deformation characteristics; and a twin prediction module that inputs the features into a three-dimensional spatial model, generates three-dimensional coordinate deviation values ​​through comparison, and determines the connection loosening parameter to output a warning command. This invention overcomes the lag inherent in relying solely on macroscopic pressure monitoring, achieving accurate early prediction and closed-loop compensation of pipeline connection conditions.

Description

Technical Field

[0001] This invention relates to the field of new energy vehicle battery thermal management and digital twin condition monitoring technology, and in particular to a pipeline connection condition prediction system based on digital twins. Background Technology

[0002] The external liquid-cooled piping of new energy vehicle battery boxes is responsible for the thermal management of the power battery. To improve system maintenance efficiency, liquid-cooled piping generally adopts quick-connect technology and has angle compensation and pressure adaptive response capabilities. Existing digital twin-based piping connection status prediction systems are deeply adapted to prefabricated and parametric designs, aiming to construct a digital model of the physical entity of the liquid-cooled piping. The digital twin model mainly extracts macroscopic fluid pressure data inside the liquid-cooled piping, combines it with fluid dynamics algorithms to analyze pressure fluctuation characteristics, and thus synchronously simulates the sealing status of physical connections and provides leakage early warning assessment.

[0003] Existing pipeline connection status prediction technologies suffer from technical pain points such as a lack of underlying microscopic sensing data and prediction lag. New energy vehicles operate under constant vibration and temperature fluctuations, making the physical connections of liquid-cooled pipelines highly susceptible to microscopic mechanical displacement and loosening. However, digital twin monitoring systems rely solely on macroscopic fluid pressure data for threshold comparisons, failing to accurately capture minute displacement changes within physical connections. This results in the digital twin model not synchronizing with the actual structural state of the liquid-cooled pipeline. For example, if a liquid-cooled pipeline connection fastener becomes loose at the millimeter level, and the internal fluid pressure has not yet reached the alarm threshold, the digital twin monitoring system will not trigger a leak warning. By the time the system senses a sudden change in macroscopic pressure and issues a warning, the physical connection has typically already experienced substantial sealing failure and coolant leakage short-circuit. The industry urgently needs innovative physical monitoring technologies to accurately predict the mechanical health status of pipeline connections in advance. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a pipeline connection status prediction system based on digital twins. This invention solves the technical problem of delayed prediction of liquid cooling pipeline connection status and sudden leakage caused by the fact that physical pipelines may develop minor loosening due to long-term vibration, which cannot be detected in real time by the digital twin system.

[0005] To solve the above-mentioned technical problems, the specific contents of the present invention are as follows: The pipeline connection status prediction system based on digital twin provided by the present invention includes equipment and control device for establishing communication connection; The device includes a liquid cooling pipeline, a fixed base, threaded rods, a conical gear, a drive disk, a servo electric drive, a displacement sensing assembly, and a rotary encoder assembly. The liquid cooling pipeline is connected to the fixed base. Two threaded rods are screwed into the interior of the fixed base. The ends of the threaded rods are connected to the conical gear. The servo electric drive is connected to the drive disk. The drive disk meshes with the conical gear. The displacement sensing assembly is located between the fixed bases. The rotary encoder assembly is mounted on the threaded rods. The control device includes a parameter acquisition module, a pulse excitation module, and a twin prediction module; The parameter acquisition module retrieves the axial displacement parameter monitored by the displacement sensing component and the rotational deflection parameter monitored by the rotational encoding component, and transmits the axial displacement parameter and the rotational deflection parameter to the pulse excitation module. The pulse excitation module receives the axial displacement parameter and the rotational deflection parameter, sets the axial displacement parameter and the rotational deflection parameter as initial reference data, sends a micro-torque pulse command to the servo electric drive, extracts the pulse response sequence corresponding to the micro-torque pulse command fed back by the servo electric drive, and transmits the pulse response sequence and the initial reference data to the twin prediction module. The twin prediction module receives the impulse response sequence and the initial reference data, inputs the impulse response sequence and the initial reference data into the three-dimensional space model, compares the initial reference data with the impulse response sequence to generate a three-dimensional coordinate deviation value, extracts the three-dimensional coordinate deviation value as a connection loosening parameter, and outputs a pipeline warning command when the connection loosening parameter exceeds a set threshold.

[0006] Furthermore, in the digital twin-based pipeline connection status prediction system of the present invention, the device further includes a temperature sensing component and a vibration sensing component; the temperature sensing component is attached to the surface of the fixing base, and the vibration sensing component is assembled on the outer wall of the liquid-cooled pipeline; the temperature sensing component collects ambient temperature alternating parameters, and the vibration sensing component collects road surface vibration frequency parameters; the temperature sensing component and the vibration sensing component transmit the ambient temperature alternating parameters and the road surface vibration frequency parameters to the parameter acquisition module.

[0007] Furthermore, in the pipeline connection status prediction system based on digital twins described in this invention, the control device includes a timing synchronization module. The timing synchronization module has a built-in clock node, extracts the output time node of the micro-torque pulse command issued by the pulse excitation module and the pulse response sequence, extracts the sampling time nodes of the axial displacement parameter, the rotational deflection parameter, the ambient temperature alternation parameter, and the road surface vibration frequency parameter retrieved by the parameter acquisition module, aligns the output time node with the sampling time node to the same moment, encapsulates the axial displacement parameter, the rotational deflection parameter, the ambient temperature alternation parameter, the road surface vibration frequency parameter, and the pulse response sequence into the same time identifier to assemble a timing synchronization parameter set, and transmits the timing synchronization parameter set to the critical verification module and the signal decoupling module.

[0008] Furthermore, in the pipeline connection status prediction system based on digital twins described in this invention, the control device includes a critical verification module. The critical verification module receives the set of timing synchronization parameters, extracts the ambient temperature alternating parameters within the set of timing synchronization parameters, retrieves a preset static friction coefficient lookup table, inputs the ambient temperature alternating parameters into the static friction coefficient lookup table to match the static friction self-locking critical value, and constrains the peak value of the torque amplitude in the micro-torque pulse command within the limited range of the static friction self-locking critical value to prevent the equipment from undergoing mechanical unlocking displacement.

[0009] Furthermore, in the pipeline connection status prediction system based on digital twins of the present invention, the control device includes a signal decoupling module; the signal decoupling module receives the time-series synchronization parameter set, extracts the ambient temperature alternating parameter and the rotational deflection parameter within the time-series synchronization parameter set, inputs the ambient temperature alternating parameter into a filtering and smoothing algorithm to calculate the thermal expansion deformation value, removes the thermal expansion deformation value from the rotational deflection parameter, and derives the corrected deflection parameter.

[0010] Furthermore, in the pipeline connection status prediction system based on digital twins described in this invention, the signal decoupling module extracts the road vibration frequency parameter from the set of time-series synchronization parameters, constructs a band-stop filter frequency band according to the road vibration frequency parameter, inputs the correction deflection parameter into the band-stop filter frequency band to perform frequency domain filtering processing, filters out specific frequency parameters, and derives the filter response sequence.

[0011] Furthermore, in the pipeline connection status prediction system based on digital twins of the present invention, the control device includes a hysteresis mapping module; the hysteresis mapping module receives the filtered response sequence, extracts the impulse response sequence within the set of time-series synchronization parameters, sets the values ​​of the impulse response sequences under the same time marker as the vertical coordinate, sets the corresponding values ​​of the filtered response sequences as the horizontal coordinate, calibrates the closed motion trajectory in two-dimensional space, uses an integral algorithm to convert the closed motion trajectory into a physical area value, and derives the physical area value and the closed motion trajectory.

[0012] Furthermore, in the digital twin-based pipeline connection state prediction system of the present invention, the control device includes a stiffness evaluation module; the stiffness evaluation module receives the physical area value and the closed motion trajectory, extracts the slope of the main diagonal of the closed motion trajectory as the dynamic contact stiffness, retrieves the built-in initial reference area value, performs a difference operation between the physical area value and the initial reference area value to obtain the area expansion rate, and derives the dynamic contact stiffness and the area expansion rate.

[0013] Furthermore, in the digital twin-based pipeline connection status prediction system of the present invention, the control device includes a life warning module; the life warning module receives the dynamic contact stiffness and the area expansion rate, inputs the area expansion rate and the dynamic contact stiffness into a pre-fitted performance degradation model, extracts the time node values ​​output by the performance degradation model, sets the time node values ​​as mechanical health margins, compares the mechanical health margins with a preset safety baseline value, and generates a life warning command when the mechanical health margins are lower than the preset safety baseline value.

[0014] Furthermore, in the pipeline connection status prediction system based on digital twins described in this invention, the control device includes a dynamic compensation module; the dynamic compensation module receives the life warning command, retrieves the mechanical health margin, and when the mechanical health margin is within a set warning value range, sends a step angle compensation operation command to the servo electric drive component, driving the servo electric drive component to drive the threaded rod to rotate.

[0015] Beneficial effects of this invention: The pipeline connection status prediction system based on digital twins provided by this invention is configured with a pulse excitation module and a twin prediction module. The pulse excitation module sends micro-torque pulse commands to the servo electric drive unit, driving the servo electric drive unit to output alternating micro-torque. Under the physical premise of not causing macroscopic mechanical unlocking displacement, the alternating micro-torque excites micro-elastic deformation in the mechanical transmission link, and the servo electric drive unit simultaneously extracts the pulse response sequence reflecting the degree of underlying micro-elastic deformation. The twin prediction module receives the pulse response sequence and initial reference data containing the static reference attitude, inputs the pulse response sequence and initial reference data into a three-dimensional spatial model for comparison and calculation, generates three-dimensional coordinate deviation values, and extracts the three-dimensional coordinate deviation values ​​as connection loosening parameters.

[0016] The underlying hardware micro-torque excitation mechanism is connected and computationally coordinated with the upper-level three-dimensional spatial model mapping mechanism. This transforms the static mechanical fastening structure into a dynamic response carrier with active dynamic feedback. Based on actively applied physical detection excitation, it accurately captures the microscopic physical attenuation characteristics of the internal contact stiffness of the bottom connection interface of the liquid-cooled pipeline. The aforementioned data acquisition and collaborative computation logic overcomes the technical shortcomings of existing pipeline connection monitoring technologies based on digital twins, which rely solely on macroscopic fluid pressure data to perform threshold comparisons. It also breaks through the technical bottleneck of severely delayed leak warning due to the lack of underlying microscopic sensing data.

[0017] The system can detect when the connection loosening parameter exceeds a set threshold and output a pipeline warning command before a substantial sealing failure occurs in the physical connection device or before the internal fluid pressure in the pipeline reaches the alarm threshold. To further coordinate with downstream control nodes, the dynamic compensation module receives the pipeline warning command and sends a step angle compensation operation command to the servo electric drive component to drive the threaded rod to replenish the lost mechanical clamping force. Ultimately, this achieves accurate early prediction and automated closed-loop intervention of the mechanical health status of physical connection nodes in liquid cooling pipelines for new energy vehicles. Attached Figure Description

[0018] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on the accompanying drawings without creative effort.

[0019] Figure 1 This is a system architecture diagram of the pipeline connection status prediction system based on digital twins according to the present invention. Detailed Implementation

[0020] To make the technical solution of the present invention clearer, the present invention will be clearly and completely described below with reference to specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. 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. The various embodiments of the present invention will be described in detail below with reference to the accompanying drawings. To better understand the purpose of the present invention, the present invention will be described in further detail below.

[0021] Please see Figure 1 The pipeline connection status prediction system based on digital twin provided by the present invention includes equipment and control device for establishing communication connection. The device includes a liquid cooling pipeline, a fixed base, threaded rods, a conical gear, a drive disk, a servo electric drive, a displacement sensing assembly, and a rotary encoder assembly. The liquid cooling pipeline is connected to the fixed base. Two threaded rods are screwed into the interior of the fixed base. The ends of the threaded rods are connected to the conical gear. The servo electric drive is connected to the drive disk. The drive disk meshes with the conical gear. The displacement sensing assembly is located between the fixed bases. The rotary encoder assembly is mounted on the threaded rods. The control device includes a parameter acquisition module, a pulse excitation module, and a twin prediction module; The parameter acquisition module retrieves the axial displacement parameter monitored by the displacement sensing component and the rotational deflection parameter monitored by the rotational encoding component, and transmits the axial displacement parameter and the rotational deflection parameter to the pulse excitation module. The pulse excitation module receives the axial displacement parameter and the rotational deflection parameter, sets the axial displacement parameter and the rotational deflection parameter as initial reference data, sends a micro-torque pulse command to the servo electric drive, extracts the pulse response sequence corresponding to the micro-torque pulse command fed back by the servo electric drive, and transmits the pulse response sequence and the initial reference data to the twin prediction module. The twin prediction module receives the impulse response sequence and the initial reference data, inputs the impulse response sequence and the initial reference data into the three-dimensional space model, compares the initial reference data with the impulse response sequence to generate a three-dimensional coordinate deviation value, extracts the three-dimensional coordinate deviation value as a connection loosening parameter, and outputs a pipeline warning command when the connection loosening parameter exceeds a set threshold.

[0022] The external liquid cooling pipeline of the battery box in new energy vehicles is responsible for transporting high-pressure coolant. Due to long-term exposure to road vibrations and alternating high and low temperatures, the physical anti-loosening structure is prone to micro-displacement. The digital twin pipeline connection status prediction system includes equipment and a control device for establishing communication connections. The equipment is deployed at the physical connection nodes of the liquid cooling pipeline, which is connected to a fixed base. Two threaded rods are screwed into the fixed base, with conical gears connected to the ends of the threaded rods. A servo electric drive unit connects to a drive disk, which meshes with the conical gears. The system applies torque through the servo electric drive unit to rotate the drive disk, which in turn drives the conical gears, causing the threaded rods to rotate in or out of the fixed base, thereby adjusting the clamping force of the fixed base on the liquid cooling pipeline. A displacement sensing component is located between the fixed bases to collect axial displacement parameters reflecting the width of the physical gap between the fixed bases; a rotational encoding component is mounted on the threaded rods to collect rotational deflection parameters mapping the torsional state of the threaded rods.

[0023] The control device incorporates a parameter acquisition module, a pulse excitation module, and a twin prediction module. The parameter acquisition module retrieves axial displacement parameters monitored by the displacement sensing component and rotational deflection parameters monitored by the rotational encoding component, and transmits these two parameters to the pulse excitation module. The pulse excitation module receives the axial displacement and rotational deflection parameters, sets the received parameters as initial reference data, which is a set of baseline three-dimensional attitude data of the mechanical structure in a static environment. Subsequently, it sends a micro-torque pulse command to the servo drive, driving the servo drive to output an alternating torque with a frequency of 100Hz and an amplitude of 0.05 N·m. The alternating torque excites micro-elastic deformation in the mechanical transmission link without causing macroscopic thread unrotation. Simultaneously, the pulse excitation module extracts the pulse response sequence fed back by the servo drive. The pulse response sequence includes time markers and corresponding electromagnetic torque values ​​during the output of the alternating torque. Finally, the pulse response sequence and the initial reference data are transmitted to the twin prediction module.

[0024] The twin prediction module receives the impulse response sequence and initial reference data, and inputs them into a pre-configured finite element mesh-based fluid-structure interaction (FSI) 3D mapping model. This FSI 3D mapping model incorporates a spatial inverse kinematics matrix. The FSI 3D mapping model extracts the electromagnetic torque value from the impulse response sequence and substitutes it into the spatial inverse kinematics matrix to convert it into a dynamic angular displacement gradient parameter. The FSI 3D mapping model retrieves the preset pitch parameter of the threaded rod and multiplies the dynamic angular displacement gradient parameter with the preset pitch parameter to obtain an equivalent axial micro-displacement value. Based on the equivalent axial micro-displacement value, the FSI 3D mapping model synchronously updates the coordinate parameters of the finite element mesh nodes representing the fixed seat in the 3D coordinate system. The FSI 3D mapping model extracts the updated finite element mesh node coordinate parameters, calculates the Euclidean distance between the updated finite element mesh node coordinate parameters and the corresponding reference node coordinate parameters in the initial reference data, and generates a 3D coordinate deviation value from the Euclidean distance. The three-dimensional coordinate deviation value is used to reflect the deformation displacement of the grid nodes of the physical fixing seat in the three-dimensional coordinate system. The twin prediction module extracts the three-dimensional coordinate deviation value as a connection loosening parameter, and outputs a pipeline warning command when it determines that the connection loosening parameter exceeds a set threshold.

[0025] The real physical environment in which the liquid cooling pipeline is located contains multiple sources of environmental interference, and the equipment is equipped with corresponding temperature and vibration sensors. The temperature sensor is attached to the surface of the mounting base and is responsible for collecting ambient temperature alternating parameters, including numerical sequences of temperature rises and falls. The vibration sensor is mounted on the outer wall of the liquid cooling pipeline and is responsible for collecting road vibration frequency parameters, including random broadband vibration values ​​generated by vehicle movement. After data acquisition, the temperature and vibration sensors transmit the ambient temperature alternating parameters and the road vibration frequency parameters to the parameter acquisition module.

[0026] The fusion of data from different physical signals is strictly dependent on the acquisition timing. The control device is equipped with a timing synchronization module, with a built-in clock node to unify the time base. The timing synchronization module extracts the output time node of the command issued by the pulse excitation module and the pulse response sequence, and at the same time extracts the sampling time nodes of parameters such as displacement, temperature, and vibration retrieved by the parameter acquisition module. Then, the output time node is aligned with the sampling time node to the same moment, and all heterogeneous parameters are encapsulated in the same time identifier to compile a timing synchronization parameter set. In this embodiment, the time identifier is specifically represented by a microsecond-level time identifier to ensure that all internal heterogeneous data have an equivalent microsecond-level time identifier record. Finally, the timing synchronization parameter set is transmitted to the critical verification module and the signal decoupling module.

[0027] Considering that different temperature environments directly affect the physical friction state of metal parts, the criticality verification module receives a set of timing synchronization parameters and extracts the ambient temperature alternation parameters within the set. The criticality verification module retrieves a preset static friction coefficient lookup table, inputs the ambient temperature alternation parameters into the lookup table, and matches the corresponding static friction self-locking critical value according to the physical mapping law. The static friction self-locking critical value is equivalent to the lower limit torque threshold that is sufficient to cause physical slippage of the threaded rod at the current temperature. Based on this, the system constrains the peak torque amplitude within the micro-torque pulse command within the limited range, preventing the risk of torque excitation exceeding the limit and triggering actual mechanical unlocking displacement.

[0028] The underlying physical parameters incorporate slow environmental thermal deformation data. The signal decoupling module receives the timing synchronization parameter set and extracts the internal environmental temperature alternation parameters and rotational deflection parameters. The module inputs the environmental temperature alternation parameters into a filtering and smoothing algorithm to calculate the thermal expansion deformation value. The thermal expansion deformation value maps to the low-frequency physical displacement change data caused by material heating. Subsequently, the thermal expansion deformation value is removed from the rotational deflection parameters, deriving a corrected deflection parameter that eliminates temperature drift errors.

[0029] The signal decoupling module continues to extract road vibration frequency parameters from the time-series synchronization parameter set, and constructs a band-stop filter band covering the random oscillation range of 1Hz to 50Hz based on the road vibration frequency parameters. The module inputs the correction deflection parameter into the band-stop filter band to perform frequency domain filtering, eliminating specific frequency parameters equivalent to road background noise data, and deriving a filter response sequence for filtering out environmental thermal vibration noise.

[0030] Physical feature mapping is performed on the filtered response sequence after noise removal. The hysteresis mapping module receives the filtered response sequence and extracts the impulse response sequence within the set of timing synchronization parameters. The hysteresis mapping module sets the impulse response sequence values ​​under the same time marker as the vertical coordinate and the corresponding filtered response sequence values ​​as the horizontal coordinate, calibrating a closed motion trajectory containing loading and unloading curves in two-dimensional space. Furthermore, the discrete trapezoidal integral algorithm is used to convert the closed motion trajectory into physical area values. The physical area values ​​are equivalent to the frictional damping energy consumed by the mechanical transmission interface in a single alternating torque cycle. Finally, the physical area values ​​and the closed motion trajectory are derived.

[0031] Given that the closed characteristic curve embodies the physical degradation law of the metal interface, the stiffness evaluation module built into the control device receives the physical area value and the closed motion trajectory, extracts the slope of the main diagonal of the closed motion trajectory as the dynamic contact stiffness, and the dynamic contact stiffness maps the actual deformation resistance of the current metal contact interface. The stiffness evaluation module retrieves the built-in initial reference area value, performs a difference operation between the physical area value and the initial reference area value to obtain the area expansion rate. The increase in the area expansion rate is equivalent to the increase in void size caused by the wear of the microstructure of the threaded engagement surface; after the evaluation is completed, the dynamic contact stiffness and area expansion rate are exported.

[0032] The lifespan early warning module receives dynamic contact stiffness and area expansion rate, and inputs these parameters into a performance degradation model with exponential decay boundary constraints. This performance degradation model is a pre-constructed mathematical fitting model based on historical vibration fatigue test data of the liquid-cooled pipeline. The control device pre-extracts a priori test sample set of the same type of liquid-cooled pipeline connectors under standard high and low temperature alternating and swept-frequency vibration destructive test conditions. This priori test sample set includes the actual contact area loss rate, contact stiffness reduction, and the time point parameter for actual sealing failure of the physical connection under different vibration durations. The control device uses the actual contact area loss rate and contact stiffness reduction as independent input variables and the time point parameter as the output dependent variable. It employs nonlinear least squares method for polynomial surface fitting calculations to calibrate the material fatigue decay constant applicable to the current pipeline material, thus constructing a performance degradation model that quantifies the mapping relationship between area expansion rate, dynamic contact stiffness, and mechanical fastening stress decay. The lifespan warning module uses a calibrated performance degradation model to extrapolate the rate of decrease in mechanical fastening stress over time. It extracts time-node values ​​from the performance degradation model, representing the remaining time window before the dynamic contact stiffness falls below the minimum working pressure of the seal. The module sets these time-node values ​​as the mechanical health margin, and generates a lifespan warning command when the mechanical health margin falls below a preset safety threshold.

[0033] To address the risk of physical structure failure, the dynamic compensation module receives a lifespan warning command and retrieves the mechanical health margin. When the mechanical health margin is determined to be within the set warning value range, the dynamic compensation module sends a step angle compensation operation command to the servo drive. Upon receiving the command, the servo drive drives the drive disk to rotate, causing the conical gear and threaded rod to rotate and generate compensating physical displacement. The servo drive drives the threaded rod to replenish the preload lost by the mechanical interface during long-term oscillation, restoring the rigid sealing and compression state between the fixed seat and the liquid cooling pipeline.

[0034] In one specific implementation, the specific data processing path and mathematical mapping relationship within the above steps are as follows: The critical verification module receives the timing synchronization parameter set and extracts the ambient temperature alternation parameter from within it. The critical verification module then retrieves a pre-set static friction coefficient lookup table. Based on the physical mapping between temperature and material expansion coefficients, the critical verification module constructs the static friction self-locking boundary equation. The static friction self-locking boundary equation is defined as follows:

[0035] Inside the formula, This represents the critical value for static friction self-locking. Represents the static friction coefficient at room temperature. The coefficient of linear expansion representing the material of the threaded rod. This represents the parameter of alternating ambient temperature. This represents the factory-set reference calibration temperature. This represents the initial mechanical preload for assembling and securing the threaded rod. The parameter represents the thread pitch diameter of the threaded rod. The critical verification module inputs the alternating ambient temperature parameter into the static friction self-locking boundary equation to calculate the critical value of static friction self-locking. The critical verification module constrains the peak torque amplitude within the micro-torque pulse command to the limited range of the static friction self-locking critical value, preventing the physical risk of the device mechanically unlocking due to torque excitation exceeding the limit. The signal decoupling module receives the time-series synchronization parameter set and extracts the alternating ambient temperature parameter and rotational deflection parameter within the time-series synchronization parameter set. The signal decoupling module constructs a first-order exponential smoothing filter equation for the time series. The first-order exponential smoothing filter equation for the time series is defined as:

[0036] Inside the formula, The value of thermal expansion deformation at the current time point. Represents the smoothing filter weight coefficients. The rotation deflection parameter represents the current time point. This represents the thermal expansion deformation value at the previous time point. The signal decoupling module maps the alternating ambient temperature parameter to smoothing filter weight coefficients. The signal decoupling module inputs the rotation deflection parameter into the first-order exponential smoothing filter equation of the time series to calculate the thermal expansion deformation value. The signal decoupling module establishes a deflection parameter correction equation. The deflection parameter correction equation is defined as:

[0037] Inside the formula, This represents the correction deflection parameter. The rotation deflection parameter represents the current time point. This represents the thermal expansion deformation value at the current time point. The signal decoupling module removes the thermal expansion deformation value from the rotational deflection parameter through the deflection parameter correction equation, deriving the corrected deflection parameter.

[0038] The hysteresis mapping module receives the filtered response sequence and extracts the impulse response sequence within the set of timing synchronization parameters. The module sets the impulse response sequence values ​​under the same time marker as the vertical coordinate and the corresponding filtered response sequence values ​​as the horizontal coordinate, thus calibrating the closed motion trajectory in two-dimensional space. The hysteresis mapping module constructs a discrete trapezoidal integral equation. The discrete trapezoidal integral equation is defined as:

[0039] Inside the formula, Represents the physical area value. Represents the total number of sampling points in the impulse response sequence. Represents the current sampling point sequence number. This represents the filtered response sequence value at time i+1. This represents the numerical value of the filtered response sequence at time i. This represents the numerical value of the impulse response sequence at time i+1. This represents the impulse response sequence value at time i. The hysteresis mapping module inputs the filtered response sequence value and the impulse response sequence value into the discrete trapezoidal integral equation, converting and calculating the physical area value enclosed by the closed motion trajectory. The hysteresis mapping module derives the physical area value and the closed motion trajectory. The stiffness evaluation module receives the physical area value and the closed motion trajectory. The stiffness evaluation module establishes the dynamic contact stiffness solution equation. The dynamic contact stiffness solution equation is defined as:

[0040] Inside the formula, Represents dynamic contact stiffness. This represents the maximum value of the impulse response sequence within a closed motion trajectory. This represents the minimum value of the impulse response sequence in a closed motion trajectory. This represents the maximum value of the filtered response sequence within the closed motion trajectory. This represents the minimum value of the filtered response sequence in the closed motion trajectory. The stiffness assessment module extracts the extreme value of the above sequence and inputs it into the dynamic contact stiffness solving equation to calculate the dynamic contact stiffness. The stiffness assessment module retrieves the built-in initial reference area value. The stiffness assessment module constructs the area expansion rate difference equation. The area expansion rate difference equation is defined as:

[0041] Inside the formula, Represents the area expansion rate. Represents the physical area value. This represents the initial reference area value. The stiffness assessment module inputs the physical area value and the initial reference area value into the area expansion rate difference equation to calculate the area expansion rate. The stiffness assessment module then exports the dynamic contact stiffness and area expansion rate.

[0042] The lifespan warning module receives dynamic contact stiffness and area expansion rate. The module constructs an exponential performance degradation equation. The exponential performance degradation equation is defined as follows:

[0043] Inside the formula, Represents the value at a specific time point. The fatigue decay constant of the material at the contact interface between the fixed seat and the threaded rod. Represents the area expansion rate. This represents the preset safety baseline value. This represents dynamic contact stiffness. The lifespan warning module inputs the area expansion rate and dynamic contact stiffness into the exponential performance degradation equation to derive the time node values ​​output by the performance degradation model. The lifespan warning module sets these time node values ​​as the mechanical health margin. The lifespan warning module compares the mechanical health margin with the preset safety baseline value to set an alarm time threshold. When the mechanical health margin falls below the alarm time threshold, the lifespan warning module generates a lifespan warning command.

[0044] The equipment is deployed at the external liquid cooling pipeline node of the battery box in a commercial new energy vehicle. The parameter acquisition module collects the alternating ambient temperature parameter of 85°C. The criticality verification module extracts the static friction coefficient at room temperature as 0.15, the linear expansion coefficient of the threaded rod material as 0.000012 / °C, the reference calibration temperature as 25°C, the initial mechanical preload as 5000N, and the thread pitch diameter parameter of the threaded rod as 0.01m. The criticality verification module substitutes the multi-dimensional environmental physical parameters into the static friction self-locking boundary equation and calculates the static friction self-locking critical value as 3.747 N·m. The control device strictly constrains the peak torque amplitude of the micro-torque pulse command to 0.5 N·m. The physical area value output by the hysteresis mapping module through the discrete trapezoidal integral equation is 0.012 J. The stiffness assessment module extracts the maximum value of the impulse response sequence from the closed motion trajectory as 0.5 N·m and the minimum value as -0.5 N·m; the maximum value of the filtered response sequence is 0.02 rad and the minimum value is -0.02 rad. The stiffness assessment module substitutes the corresponding extreme values ​​into the dynamic contact stiffness equation, calculating a dynamic contact stiffness of 25 N·m / rad. The initial reference area is preset to 0.010 J, and the stiffness assessment module substitutes this into the area expansion rate difference equation, calculating an area expansion rate of 20%. The lifespan warning module configures the material fatigue decay constant to 0.005 and extracts a preset safety baseline value of 15 N·m / rad. The lifespan warning module substitutes the dynamic contact stiffness of 25 N·m / rad and the area expansion rate of 20% into the exponential performance degradation equation, calculating a time node value of 510 h. The mechanical health margin is simultaneously calibrated to 510 h. The lifespan warning module sets the alarm time threshold to 500 h. The current mechanical health margin is higher than the alarm time threshold, and the pipeline connection status remains within the safe range. Due to the long-term effects of road surface vibrations, the mechanical health margin calculated in subsequent detection cycles has decreased to below 490 hours. The lifespan warning module determines that the value exceeds the alarm time threshold and immediately generates a lifespan warning command. The dynamic compensation module receives the lifespan warning command and sends a step angle compensation operation command to the downstream servo electric drive components.

[0045] In another preferred embodiment, to achieve the physical intervention closed loop of the above-mentioned control device, the hardware coordination and instruction scheduling architecture of this system is as follows: The physical connection nodes of the external liquid cooling pipelines in commercial new energy vehicle battery boxes are constantly exposed to high-frequency mechanical stress impacts caused by road bumps. To address this complex vibration environment, the main control board of the control device is equipped with a 32-bit reduced instruction set microprocessor (RISC) unit. Internally, the microprocessor unit integrates a high-speed serial peripheral interface bus controller, a control area network bus controller, and a direct memory access controller. The displacement sensing component and rotation encoder component integrate a high-precision analog-to-digital converter (ADC) front-end, establishing a physical signal transmission channel with the microprocessor unit via the high-speed serial peripheral interface bus. The microprocessor unit is equipped with an independent high-frequency crystal oscillator, providing a 120 MHz operating frequency to the global hardware bus to meet the high-frequency monitoring requirements of the liquid cooling pipeline connection status.

[0046] Data transmission of the parameter acquisition module is performed through the direct memory access controller. The microprocessor unit's built-in hardware timer overflows and triggers a periodic data capture interrupt. The underlying firmware responds to the data bus interrupt and triggers the high-speed serial peripheral interface bus controller. The bus controller initiates a hardware addressing operation on the internal data registers of the displacement sensing component, and the microprocessor unit sets the bus addressing parameters to a preset physical address space. The direct memory access controller acquires control of the data transmission, bypassing the microprocessor unit's arithmetic logic unit, and directly writes the read axial displacement and rotational deflection parameters into a contiguous address segment defined by the static random access memory, completing the low-latency transfer of the underlying sensing data.

[0047] The pulse excitation module writes micro-torque pulse commands to the transmit mailbox of the control area network bus controller via the microprocessor unit. These micro-torque pulse commands are encapsulated into standard data frames with specific frame identifiers. The bus controller pushes the standard data frames to the external communication bus. The servo drive's built-in digital signal processor receives and parses the standard data frames, and the underlying hardware state machine immediately switches from passive monitoring to torque output. The digital signal processor controls the internal inverter bridge to output a pulse-width modulated voltage waveform, driving the servo drive to rotate the drive disk according to the pulse-width modulated voltage waveform. The drive disk meshes with the conical gear, activating the underlying mechanical link without causing macroscopic loosening of the liquid cooling piping.

[0048] The servo drive's internal current sampling circuit synchronously generates a pulse response sequence containing stator current parameters and transmits this sequence back to the control device via the control area network bus. The timing synchronization module utilizes the microprocessor's built-in input capture-compare register (ICC) for hardware-level clock alignment, with the ICC having a 32-bit register width. When the microprocessor pin detects a level transition at the start of the pulse response sequence, the hardware triggers a pulse capture action, and the current count value of the hardware timer is immediately latched into the ICC. The microprocessor uses this latched count value as a unified time identifier, attaching it to the axial displacement parameter, rotational deflection parameter, and the frame header of the pulse response sequence, achieving microsecond-level timing binding of heterogeneous parameters.

[0049] The twin prediction module relies on the hardware coprocessor inside the control device to perform state determination. When the hardware comparator triggers a low-level safety threshold interrupt, the hardware coprocessor sends a high-priority exception interrupt request to the microprocessor unit kernel. The microprocessor unit kernel suspends the background polling task and writes the hexadecimal control machine code of the step angle compensation instruction to the control register of the servo drive via the data bus. The digital signal processor of the servo drive reads the hexadecimal control machine code, executes the low-level state machine compensation jump action, and outputs a power signal with a preset duty cycle. The servo drive further drives the threaded rod to physically advance inside the fixed seat, using the physical advance action to compensate for the clamping force between the liquid cooling pipeline and the fixed seat. Through the above steps, the control device completes the closed loop of state perception and physical intervention for the hardware level of the liquid cooling pipeline in new energy vehicles.

Claims

1. A pipeline connection status prediction system based on digital twins, characterized in that, This includes equipment and control devices for establishing communication connections; The device includes a liquid cooling pipeline, a fixed base, threaded rods, a conical gear, a drive disk, a servo electric drive, a displacement sensing assembly, and a rotary encoder assembly. The liquid cooling pipeline is connected to the fixed base. Two threaded rods are screwed into the interior of the fixed base. The ends of the threaded rods are connected to the conical gear. The servo electric drive is connected to the drive disk. The drive disk meshes with the conical gear. The displacement sensing assembly is located between the fixed bases. The rotary encoder assembly is mounted on the threaded rods. The control device includes a parameter acquisition module, a pulse excitation module, and a twin prediction module; The parameter acquisition module retrieves the axial displacement parameter monitored by the displacement sensing component and the rotational deflection parameter monitored by the rotational encoding component, and transmits the axial displacement parameter and the rotational deflection parameter to the pulse excitation module. The pulse excitation module receives the axial displacement parameter and the rotational deflection parameter, sets the axial displacement parameter and the rotational deflection parameter as initial reference data, sends a micro-torque pulse command to the servo electric drive, extracts the pulse response sequence corresponding to the micro-torque pulse command fed back by the servo electric drive, and transmits the pulse response sequence and the initial reference data to the twin prediction module. The twin prediction module receives the impulse response sequence and the initial reference data, inputs the impulse response sequence and the initial reference data into the three-dimensional space model, compares the initial reference data with the impulse response sequence to generate a three-dimensional coordinate deviation value, extracts the three-dimensional coordinate deviation value as a connection loosening parameter, and outputs a pipeline warning command when the connection loosening parameter exceeds a set threshold.

2. The digital-twin-based pipe connection state prediction system of claim 1, wherein, The device further includes a temperature sensing component and a vibration sensing component; the temperature sensing component is attached to the surface of the mounting base, and the vibration sensing component is mounted on the outer wall of the liquid cooling pipeline; the temperature sensing component collects ambient temperature alternating parameters, and the vibration sensing component collects road surface vibration frequency parameters; the temperature sensing component and the vibration sensing component transmit the ambient temperature alternating parameters and the road surface vibration frequency parameters to the parameter acquisition module.

3. The digital twin-based pipe connection state prediction system of claim 2, wherein, The control device includes a timing synchronization module. The timing synchronization module has a built-in clock node, extracts the output time node of the micro-torque pulse command issued by the pulse excitation module and the pulse response sequence, extracts the sampling time nodes of the axial displacement parameter, the rotational deflection parameter, the ambient temperature alternation parameter and the road vibration frequency parameter retrieved by the parameter acquisition module, aligns the output time node with the sampling time node to the same time, encapsulates the axial displacement parameter, the rotational deflection parameter, the ambient temperature alternation parameter, the road vibration frequency parameter and the pulse response sequence into the same time identifier to assemble a timing synchronization parameter set, and transmits the timing synchronization parameter set to the critical verification module and the signal decoupling module.

4. The digital-twin-based pipe connection state prediction system of claim 3, wherein, The control device includes a critical verification module; the critical verification module receives the timing synchronization parameter set, extracts the ambient temperature alternating parameter within the timing synchronization parameter set, retrieves a preset static friction coefficient lookup table, inputs the ambient temperature alternating parameter into the static friction coefficient lookup table to match the static friction self-locking critical value, and constrains the peak value of the torque amplitude in the micro-torque pulse command within the limited range of the static friction self-locking critical value to prevent the device from mechanically unlocking or displacing.

5. The digital twin-based pipe connection state prediction system of claim 4, wherein, The control device includes a signal decoupling module; the signal decoupling module receives the timing synchronization parameter set, extracts the ambient temperature alternating parameter and the rotational deflection parameter inside the timing synchronization parameter set, inputs the ambient temperature alternating parameter into a filtering and smoothing algorithm to calculate the thermal expansion deformation value, removes the thermal expansion deformation value from the rotational deflection parameter, and derives the corrected deflection parameter.

6. The digital twin-based pipe connection state prediction system of claim 5, wherein, The signal decoupling module extracts the road vibration frequency parameters from the set of timing synchronization parameters, constructs a band-stop filter band according to the road vibration frequency parameters, inputs the correction deflection parameters into the band-stop filter band to perform frequency domain filtering, filters out specific frequency parameters, and derives the filter response sequence.

7. The digital twin-based pipe connection state prediction system of claim 6, wherein, The control device includes a hysteresis mapping module; the hysteresis mapping module receives the filtered response sequence, extracts the impulse response sequence within the set of timing synchronization parameters, sets the values ​​of the impulse response sequences under the same time marker as the vertical coordinate, sets the corresponding values ​​of the filtered response sequences as the horizontal coordinate, calibrates the closed motion trajectory in two-dimensional space, uses an integral algorithm to convert the closed motion trajectory into a physical area value, and derives the physical area value and the closed motion trajectory.

8. The digital-twin-based pipe connection state prediction system of claim 7, wherein, The control device includes a stiffness evaluation module; the stiffness evaluation module receives the physical area value and the closed motion trajectory, extracts the slope of the main diagonal of the closed motion trajectory as the dynamic contact stiffness, retrieves the built-in initial reference area value, performs a difference operation between the physical area value and the initial reference area value to obtain the area expansion rate, and derives the dynamic contact stiffness and the area expansion rate.

9. The digital twin-based pipe connection state prediction system of claim 8, wherein, The control device includes a lifespan warning module; the lifespan warning module receives the dynamic contact stiffness and the area expansion rate, inputs the area expansion rate and the dynamic contact stiffness into a pre-fitted performance degradation model, extracts the time node values ​​output by the performance degradation model, sets the time node values ​​as mechanical health margins, compares the mechanical health margins with preset safety baseline values, and generates a lifespan warning command when the mechanical health margins are lower than the preset safety baseline values.

10. The digital twin-based pipe connection state prediction system of claim 9, wherein, The control device includes a dynamic compensation module; the dynamic compensation module receives the life warning command, retrieves the mechanical health margin, and when the mechanical health margin is within the set warning value range, sends a step angle compensation operation command to the servo electric drive to drive the threaded rod to rotate.

Images