Magnetic performance dynamic compensation method of magnetic latching relay adapting to wide temperature range environment

Abstract

The application discloses a magnetic performance dynamic compensation method of a magnetic latching relay suitable for a wide temperature range environment, relates to the field of magnetic performance dynamic compensation, and acquires real-time temperature data and real-time load voltage data of the relay, outputs basic driving energy parameters, constructs a three-dimensional stiffness potential energy surface dimension reduction solution equivalent stiffness coefficient based on the real-time temperature data, predicts mechanical collision kinetic energy at the moment when a moving contact is closed, generates a critical rebound energy threshold, constructs an ideal convergence trajectory satisfying physical deceleration constraints, and performs inverse dynamics solution to generate a composite step driving waveform instruction, drives the relay coil by executing the composite step driving waveform instruction, solves the motion state vector of a moving iron core in real time through a flux linkage observer, identifies mechanical rebound occurrence and triggers active magnetic damping adjustment, forcibly clamps the moving iron core in a magnetic latching balance area, forcibly locks the moving contact in a conduction position, and completely avoids false attraction failure.

Description

Technical Field

[0001] This invention relates to the field of dynamic magnetic performance compensation, specifically a method for dynamic magnetic performance compensation of magnetic latching relays adapted to a wide temperature range environment. Background Technology

[0002] Existing magnetic latching relay drive control technologies, especially in new energy vehicles or smart grid applications certified for wide temperature ranges (e.g., -40℃ to 85℃), generally employ open-loop overdrive compensation strategies based on the "worst operating condition boundary." To overcome the difficulty of engagement when coil resistance decreases but mechanical impedance increases at low temperatures, or when the supply voltage drops to the lower limit, existing technologies typically incorporate a static temperature-voltage-pulse width three-dimensional lookup table. The core logic is to ensure that the electromagnetic ampere-turns are sufficient to overcome the maximum static friction and the preload of the reaction spring by unidirectionally increasing the drive energy. This strategy essentially simplifies the nonlinear electromagnetic-mechanical coupling system into a static load model, focusing on ensuring the lower limit of energy required for engagement, while neglecting the real-time matching of input energy with the dynamic mechanical stiffness characteristics of the contact spring at different temperature points.

[0003] However, this unidirectional energy gain mechanism, designed to eliminate "insufficient engagement," is highly susceptible to inducing a hidden "secondary rebound disconnection" fault under specific temperature and pressure conditions. Specifically, when the ambient temperature is within a sensitive range where the elastic modulus of the contact spring material transitions from a cold, brittle state to a highly elastic state, such as around 0°C to 25°C, and the coil supply voltage is not at an extremely low value, the excessive compensation energy based on the low temperature or low voltage preset cannot be completely consumed by mechanical work and is instead converted into excessive end impact kinetic energy of the moving iron core. Because existing control methods fail to detect the nonlinear stiffness fluctuations caused by thermally induced lattice distortion of the spring material, the moving contact experiences a violent rigid collision at the moment of closure. The instantaneous reverse mechanical rebound acceleration and rebound torque generated momentarily exceed the magnetic holding torque of the permanent magnet, which is not yet fully stable. This energy mismatch caused by overcompensation results in the moving iron core rebounding and tripping within milliseconds after physical closure, forming a false "engagement-instantaneous disconnection" action. Conventional low-frequency state detection often struggles to capture such transient physical failures caused by mechanical overshoot.

[0004] To address the aforementioned shortcomings, a technical solution is provided. Summary of the Invention

[0005] To address the technical problems mentioned in the background section, this invention is proposed. This invention provides a method for dynamically compensating the magnetic properties of a magnetic latching relay adapted to a wide temperature range environment.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A method for dynamic compensation of magnetic properties of magnetic latching relays adapted to a wide temperature range includes the following steps:

[0008] Acquire real-time temperature data of the current working environment of the relay and real-time load voltage data of the power supply, construct an initial magnetic circuit drive model based on temperature-voltage dual variables, and output basic drive energy parameters.

[0009] Based on the real-time temperature data, a three-dimensional stiffness potential energy surface is constructed to calculate the equivalent stiffness coefficient. Combined with the basic driving energy parameters, the mechanical collision kinetic energy at the moment of closing of the moving contact is predicted, and the critical rebound energy threshold is generated.

[0010] Based on the critical rebound energy threshold, an ideal convergence trajectory that satisfies the physical deceleration constraint is constructed, and inverse dynamics solution is performed to generate a composite step drive waveform command.

[0011] The composite stepped drive waveform command is executed to drive the relay coil. The terminal voltage and current response of the coil are collected in the drive gap window, and the motion state vector of the moving iron core is calculated in real time through the flux linkage observer.

[0012] Based on the motion state vector, mechanical rebound is detected and active magnetic damping adjustment is triggered to forcibly clamp the moving iron core within the magnetically maintained balance zone.

[0013] Furthermore, the analytical steps for constructing the three-dimensional stiffness potential energy surface and calculating the equivalent stiffness coefficient are as follows:

[0014] Based on the thermal drift vector, the thermal solid mesh is reconstructed, and the principal stress transfer skeleton is extracted through topology trimming to construct a three-dimensional stiffness potential energy surface.

[0015] The equivalent stiffness coefficient is calculated by using real-time temperature to extract dimensionality reduction features from the three-dimensional stiffness potential energy surface.

[0016] Furthermore, the analytical steps for calculating the equivalent stiffness coefficient are as follows:

[0017] Based on the direction of motion of the moving iron core, the stiffness contribution analysis and topology trimming are performed on the hot solid mesh to extract the main stress transmission skeleton connecting the fixed end and the contact end.

[0018] The bending deformation process of the principal stress transmission skeleton along the motion trajectory is mapped as a three-dimensional stiffness potential energy surface, the dimensions of which include contact point displacement, lattice distortion degree, and accumulated elastic potential energy.

[0019] Furthermore, the analytical step of calculating the equivalent stiffness coefficient also includes:

[0020] Using real-time ambient temperature data as a slicing plane, the three-dimensional stiffness potential energy surface is dimensionally reduced and cut to obtain force and displacement characteristic curves.

[0021] Geometric differentiation is performed on the force-displacement characteristic curve to calculate the slope of the tangent line near the preset operating point, thereby obtaining the equivalent stiffness flux.

[0022] The equivalent stiffness flux is scalarized, and the equivalent stiffness coefficient is output.

[0023] Furthermore, the analytical steps for generating the critical rebound energy threshold are as follows:

[0024] Based on the basic driving energy parameters and the load voltage data, an electromagnetic conversion efficiency function is constructed to calculate the theoretical impact energy.

[0025] Substitute the equivalent stiffness coefficient and theoretical impact energy into the collision recovery coefficient model to calculate the expected rebound potential energy after the collision;

[0026] If the expected rebound potential energy after the collision is greater than the minimum attraction barrier, the current state is marked as a high-risk rebound zone, and the critical rebound energy threshold is output.

[0027] Furthermore, the analytical steps for constructing the ideal convergence trajectory that satisfies the physical deceleration constraint are as follows:

[0028] A displacement-velocity phase plane of the moving iron core is constructed. Using the time-domain parameterized mapping method, combined with the initial disconnection state, the contact just-contact state, and the soft landing velocity condition, a set of motion equations for the moving iron core is constructed to generate an ideal convergent trajectory.

[0029] Furthermore, the analysis steps for performing inverse dynamics solving to generate the composite step drive waveform command are as follows:

[0030] The motion stroke of the moving iron core is discretized, and the target kinetic energy increment within each discrete displacement step is determined based on the ideal convergence trajectory.

[0031] Based on the target kinetic energy increment, a work-energy balance equation is constructed, the net electromagnetic force is calculated, and the duty cycle control sequence of the driving pulse is obtained by combining the current load voltage and the real-time coil resistance.

[0032] The duty cycle control sequence is discretized in the time domain to generate the composite step drive waveform instruction, which includes a high-energy acceleration in the front section, inertial coasting in the middle section, and PWM chopping speed limiting in the final section.

[0033] Furthermore, the analysis steps for real-time calculation of the motion state vector of the moving iron core using the flux linkage observer are as follows:

[0034] Establish the dynamic voltage balance equation of the relay, and use the collected terminal voltage and current response to separate the coil resistance voltage drop and inductance voltage drop, and extract the induced back electromotive force component.

[0035] The extended Kalman filter algorithm is used to suppress noise and estimate the state of the induced back electromotive force component. The corrected coil resistance value is obtained by estimating the coil resistance drift error in real time.

[0036] Furthermore, the analysis step of calculating the motion state vector of the moving iron core in real time through the flux linkage observer also includes:

[0037] The voltage equation is integrated based on the corrected coil resistance value to obtain the real-time flux linkage value. The real-time air gap distance and instantaneous velocity of the moving iron core relative to the stationary contact are calculated based on the real-time flux linkage value, and the motion state vector is generated by combining them.

[0038] Furthermore, the analytical steps for forcibly clamping the moving iron core within the magnetically maintained balance region are as follows:

[0039] The velocity component in the motion state vector is monitored in real time. When the velocity sign changes from positive to negative and the displacement is within the closed interval of the contact point, it is determined that a mechanical rebound has occurred, and the rebound acceleration is calculated.

[0040] Based on the rebound acceleration and the mass of the moving iron core, the reverse Lorentz force required to counteract the rebound momentum is calculated, and the required damping current amplitude is calculated based on the current magnetic circuit saturation.

[0041] The control H-bridge drive circuit flips the voltage polarity within a preset microsecond time window, pumps the coil current to the damping current amplitude, and forces the moving iron core to clamp in the magnetic holding balance area, thus completing the closed-loop correction of the attraction action.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] This invention acquires real-time temperature data of the relay's current operating environment and real-time load voltage data of the power supply, constructs an initial magnetic circuit drive model based on a temperature-voltage dual-variable model, outputs basic drive energy parameters, and, based on the real-time temperature data, constructs a three-dimensional stiffness potential energy surface to calculate the equivalent stiffness coefficient. Combined with the basic drive energy parameters, it predicts the mechanical collision kinetic energy at the instant the moving contact closes, generates a critical rebound energy threshold, and, based on this critical rebound energy threshold, constructs an ideal convergence trajectory that satisfies physical deceleration constraints. It then performs inverse dynamics solving to generate a composite stepped drive waveform command. Based on the extraction of the three-dimensional stiffness potential energy surface and topological skeleton using thermally induced lattice distortion, the nonlinear stiffness fluctuations of the moving spring caused by material modulus evolution in a specific temperature range are precisely quantified at the microscopic physical level. By mapping this dynamic stiffness characteristic to a critical rebound energy threshold and planning an ideal convergence trajectory including a final PWM chopping speed limit, the system can actively reduce the excess kinetic energy of the moving iron core at the closing end, forcing it to softly land at a low speed. This on-demand energy management analysis effectively solves the technical problem of existing technologies causing severe rigid collisions and mechanical overshoots induced by blindly applying excessive driving work in the high elastic temperature range of materials.

[0044] This invention drives a relay coil by executing a composite stepped drive waveform command, and collects the coil's terminal voltage and current response during the drive gap window. A flux linkage observer is used to calculate the motion state vector of the moving iron core in real time. Based on this motion state vector, mechanical rebound is detected, triggering active magnetic damping adjustment to forcibly clamp the moving iron core within the magnetically held equilibrium region. To address transient tripping phenomena that may occur after mechanical collisions, this invention establishes a flux linkage observer based on extended Kalman filtering, reconstructing the true displacement and velocity vector of the moving iron core within an invisible magnetic field in real time without the need for external sensors. Once the moving iron core's velocity reverses and the rebound kinetic energy exceeds the magnetic potential well boundary, the system immediately triggers a microsecond-level active magnetic damping suppression mechanism. By injecting a precisely reverse current, corrected for magnetic saturation nonlinearity, into the coil, a strong Lorentz clamping force is generated that strictly counteracts the rebound momentum. This closed-loop correction method, transforming passive holding into active clamping, ensures that the relay can forcibly lock the moving contact in the conducting position even under the most severe wide-temperature voltage fluctuation conditions, completely avoiding false engagement faults. Attached Figure Description

[0045] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. The following drawings are not drawn to scale according to the actual size, but are intended to show the main idea of ​​the present invention.

[0046] Figure 1 Flowchart of a method for dynamic compensation of magnetic properties of magnetic latching relays adapted to wide temperature range environments;

[0047] Figure 2 The flowchart shows the process of calculating the equivalent stiffness coefficient.

[0048] Figure 3 A schematic diagram illustrating the principle of dimensionality reduction and mechanical feature extraction of a three-dimensional stiffness potential energy surface;

[0049] Figure 4 This is a schematic diagram of thermal drift and skeletal layering. Detailed Implementation

[0050] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are also within the scope of protection of the present invention.

[0051] like Figure 1 As shown, the method for dynamic compensation of magnetic properties of a magnetic latching relay adapted to a wide temperature range environment includes the following steps:

[0052] Step S100: Obtain real-time temperature data of the current working environment of the relay and real-time load voltage data of the power supply, construct an initial magnetic circuit drive model based on temperature-voltage dual variables, and output basic drive energy parameters.

[0053] Real-time ambient temperature is collected using a high-precision NTC thermistor closely attached to the coil. Real-time load voltage is monitored using a microcontroller ADC module. Based on the physical properties of copper wire, a function of coil resistance changing with temperature is constructed. ,in The nominal resistance at 20℃ Specifically referring to the temperature coefficient of resistance of copper wire, with a value of 0.00393 / ℃, this parameter is obtained by consulting the material handbook and is used to accurately calculate the current coil impedance. The method for constructing the initial magnetic circuit drive model is as follows: During the offline calibration phase, a relay test platform is built, traversing preset temperature and voltage ranges, such as -40℃ to 85℃ for the temperature range and 85% to 110% of the rated voltage for the voltage range. At each specific temperature-voltage combination point, a drive signal with gradually increasing pulse width is applied to the relay, and the contact state is monitored. A reliable engaging action is defined as the moving iron core being able to completely overcome the counterforce spring and maintain a non-shaking closed contact state after contact. A non-shaking closed contact state with a non-contact resistance consistently less than a preset value, such as 10mΩ, is recorded, and the corresponding minimum pulse width time is recorded. This serves as the minimum excitation energy characterization value under this operating condition, thus forming a value indexed by temperature and voltage, with the output value being... A two-dimensional lookup table, stored in non-volatile memory, constitutes the initial magnetic circuit drive model. During system operation, data collected in real time will be... and Substitute the values ​​into the lookup table and calculate the current base pulse width time using a bilinear interpolation algorithm. This parameter represents the theoretical minimum drive time without considering dynamic wear.

[0054] The method for dynamic compensation of magnetic properties of magnetic latching relays adapted to wide temperature range environments also includes the following steps:

[0055] Step S200: Based on the real-time temperature data, construct a three-dimensional stiffness potential energy surface to calculate the equivalent stiffness coefficient, and combine it with the basic driving energy parameters to predict the mechanical collision kinetic energy at the moment of closing of the moving contact, and generate the critical rebound energy threshold.

[0056] Step S200 involves inputting the real-time temperature data into a preset thermo-mechanical coupling stiffness model of the contact spring, specifically including the following steps:

[0057] Step S201: Reconstruct the thermal solid mesh based on the thermally induced drift vector, extract the principal stress transfer skeleton through topology trimming, construct a three-dimensional stiffness potential energy surface, and use real-time temperature to perform dimensionality reduction feature extraction to calculate the equivalent stiffness coefficient.

[0058] Step S2011: Discretize the relay moving spring into a lattice topological mesh, superimpose the thermal drift vector calculated based on the real-time ambient temperature, and reconstruct the thermal solid mesh that has undergone microscopic deformation;

[0059] The CAD 3D geometric model of the moving spring is imported into the finite element preprocessing unit. Depending on the complexity of the geometry, hexahedral or tetrahedral elements are selected to mesh the model, thereby generating a mesh containing... Each node and The initial lattice topology of each unit is defined as follows: hexahedral units are regular hexahedral topologies, and tetrahedral units are triangular pyramidal topologies. To accurately simulate the physical expansion behavior of the moving spring ("fixed end constraint, free end extension"), a local Lagrange coordinate system is established with the center of the root of the fixed end of the moving spring as the origin (0,0,0). In this coordinate system, the extension direction along the length of the moving spring from the fixed end to the free end is defined as the X-axis; the direction along the width of the moving spring is defined as the Y-axis; the direction perpendicular to the surface of the moving spring, corresponding to the trajectory of the moving iron core's attraction, is defined as the Z-axis; and the direction in which the moving iron core moves away from the stationary iron core, i.e., the release direction, is defined as the positive Z-axis. The first unit is recorded in this coordinate system. Initial spatial coordinates of each node Based on pre-stored reed material properties, obtain the thermal expansion coefficient function corresponding to the current material. T represents the temperature variable. This is combined with the real-time ambient temperature obtained in step S100. With the preset reference temperature The standard room temperature is usually set at 20°C, and the temperature difference between the two is calculated. Based on the principle of linear thermal expansion, the calculation is performed for each node. Thermal drift vector relative to the fixed origin The calculation formula is: , Indicates the first The initial spatial coordinates of each node are its position vectors relative to the origin (0,0,0). This represents the coefficient of thermal expansion at real-time ambient temperature. The physical meaning of this formula is that the microscopic displacement of a material is proportional to the temperature difference, and the cumulative expansion displacement is greater for nodes farther from the fixed origin. Using the principle of vector superposition, the calculated thermally induced drift vector is superimposed onto the initial coordinates of each node, thus performing an update operation. Traverse all After each node completes its coordinate update, a thermal solid mesh that accurately reflects the dimensional elongation and morphological changes that have occurred at the current temperature is reconstructed in the virtual environment. This thermal solid mesh will serve as the geometric reference for accurate stiffness analysis in subsequent steps to eliminate simulation errors caused by neglecting thermal expansion deformation.

[0060] Step S2011 discretizes the relay reed into a lattice topological mesh, based on the fundamental theories of continuum mechanics and finite element analysis. This theory states that any complex continuous elastic body can be mathematically equivalent to a discrete set composed of a finite number of nodes and elements. By solving for the displacements of the discrete nodes, the actual deformation behavior of the continuum can be approximated. This is a recognized numerical calculation standard in engineering mechanics. The implementation of superimposing thermally induced drift vectors to reconstruct the thermal solid mesh strictly follows the linear thermal expansion law and Lagrange description method in solid-state physics. Since the thermal expansion coefficient of a material is an objective physical property, any material element will inevitably produce a definite spatial displacement under temperature difference. This displacement vector... The distance from the fixed end is proportional to the temperature difference. Therefore, updating the node coordinates through vector superposition is essentially a high-fidelity mathematical mapping and geometric reconstruction of the thermal deformation behavior of the physical entity in virtual space. This approach makes the subsequent stiffness analysis no longer based on the ideal room temperature geometric model, but on the real thermal model that has undergone micro-dimensional elongation and morphological distortion. This allows for the accurate capture of the structural geometric stiffening or softening effect caused by thermal expansion, effectively solving the technical problem of stiffness prediction deviation and rebound suppression failure caused by neglecting the micro-influence of temperature on the geometric configuration of the reed in existing technologies.

[0061] Step S2012: Based on the direction of motion of the moving iron core, perform stiffness contribution analysis and topology trimming on the hot solid mesh to extract the main stress transfer skeleton connecting the fixed end and the contact end;

[0062] like Figure 4 The diagram shown illustrates thermal drift and skeletal layering. (Refer to...) Figure 4 This paper presents a mechanical analysis model of a moving spring under thermal conditions. The bottom dashed box represents the initial position of the moving spring, i.e., its geometric contour at room temperature or under no stress. The gray-filled area in the middle represents the extracted principal stress transfer skeleton. The blank area above represents the low-energy single-crystal material identified as an invalid dead zone. The black solid arrow pointing downwards on the right represents the applied unit virtual load. The thin arrow pointing upwards in the upper right corner represents the thermally induced drift vector of the node relative to its initial position. The slanted shaded area on the left represents the fixed end constraint.

[0063] Reference Figure 4 This description uses the physical state of a relay's moving spring after heating and stress as an example. The main stress transmission frame is usually located on the core path connecting the fixed end and the contact end. Figure 4 The area is marked with a gray fill. It simulates the actual physical installation state on a hot solid mesh. (Refer to...) Figure 4 On the left, select all nodes in the mesh belonging to the fixed-end region, and force all six degrees of freedom of these nodes to zero, implementing full constraint settings. (Refer to...) Figure 4 On the right side, a unit virtual load is applied to the contact end of the moving reed. For example, 1N, the direction is along the direction of the moving iron core's attraction movement, i.e., the negative direction of the Z-axis. The reason for setting this direction is that when the relay is working, the main deformation of the moving spring is caused by the electromagnetic attraction generated by the moving iron core. The direction of this attraction is determined, and the skeleton that needs to be identified is the material path that exists to resist the deformation in this specific direction. Therefore, a detection force must be applied along this actual force direction in order to stimulate the most accurate force transmission path.

[0064] The solver traverses every volume element in the mesh. A tetrahedral or hexahedral micro-element formed by several nodes. Representative unit, subscript The element sequence index is used by the solver to traverse the mesh in order to obtain the element stiffness matrix. Construct the strain-displacement matrix This process introduces the shape function of the unit. To establish a local natural coordinate system To the global physical coordinate system The mapping, using shape functions as mathematical tools, is essentially a set of interpolation basis functions, a well-known technique in the field. For linear elements, trilinear interpolation polynomials are typically used. Their core function is to establish a bridge, mapping the standard local natural coordinate system (a regular, dimensionless ideal element) in computational space to the global physical coordinate system in real space, i.e., the actual, potentially distorted geometry of the mesh under thermal conditions. Through shape functions, the physical coordinates of any point within an element can be obtained by interpolating the coordinates of all nodes within that element; utilizing the current thermal coordinates of the nodes... Constructing the Jacobian matrix ,in , , These represent the actual coordinates of the mesh nodes in the global coordinate system after thermal expansion and deformation. , , Coordinate values. Jacobian matrix. Each element is a partial derivative of the global coordinates with respect to the local coordinates, for example... thermal coordinates The value directly determines Regarding the content of the matrix, it should be noted that the method for constructing the Jacobian matrix [J] follows the standard coordinate transformation theory in finite element analysis, which is a well-known existing technique. The characteristic of this embodiment lies in substituting the thermal coordinates into it for calculation, rather than changing the matrix structure itself. Solver calculation inverse matrix The chain rule is used to transform the derivative of the shape function with respect to local coordinates into its derivative with respect to global coordinates, i.e. Etc., explicitly construct the strain-displacement matrix. The derivation of the strain-displacement matrix [B] and the composition of its internal partial derivatives are based on general algorithms in continuum mechanics, which are conventional techniques in this field. Therefore, the complex full structure of the matrix will not be detailed here. The material constitutive matrix, determined by Young's modulus and Poisson's ratio, is then considered. By volume integral The element stiffness matrix is ​​calculated, and the thermal coordinates determine the Jacobian matrix. and its determinant Directly controlled Matrix and final stiffness matrix The values ​​ensure that the geometric nonlinearity caused by thermal expansion is accurately accounted for. Representing the strain-displacement matrix The transpose of . Matrix form based on Hooke's Law. ,in For the applied virtual load vector, Representing the overall stiffness matrix, we iterate through all elements in the mesh and calculate the stiffness matrix of each element. By accumulating the results, the system of linear equations is solved using numerical linear algebra methods, such as the conjugate gradient method or the direct method, to obtain the small elastic displacement vector of each node in the mesh after being subjected to force. For each unit Using the solved nodal displacements And the aforementioned strain-displacement matrix Calculate the strain tensor inside the element. According to the formula The elastic potential energy density stored within the unit volume was calculated. ,in, This represents the displacement of all nodes in the field.

[0065] Find the maximum strain energy density value in the entire field. Define stiffness contribution This value intuitively quantifies the contribution of each micro-element to resisting overall bending: a larger value indicates that the element has deformed and stored a large amount of energy, making it the core path of stress; a smaller value indicates that the element has hardly deformed and is in a stress blind zone. Topology trimming is performed to extract the skeleton, specifically by setting a stiffness contribution threshold. For example, taking 0.1, iterate through all units. Low-energy elements are marked as invalid dead zones and removed from the computational model, which is physically equivalent to cutting off edge material that does not bear the load. After removal, the remaining elements satisfy... High strain energy density elements, because they are the necessary pathways for force transmission from the point of application to the fixed point in their physical structure, naturally maintain a continuous spatial distribution, thus forming a topological path from the fixed end to the contact end. This path is defined as the principal stress transmission skeleton, i.e. Figure 4 The gray area, generated based on a thermal grid, accurately reflects the true mechanical transmission characteristics of the reed at the current temperature.

[0066] Step S2012 performs stiffness contribution analysis and extracts the principal stress transmission skeleton, which originates from the energy principle and force transmission path theory in elasticity. That is, during the deformation of an elastic body under stress, the work done by the external force is converted into strain energy inside the object. Moreover, the energy is not uniformly distributed, but is concentrated and transmitted along the path with the greatest material stiffness (i.e., the principal stress line). By calculating the strain energy density of each micro-grid element, the system objectively quantifies the "contribution weight" of the micro-element to the overall deformation resistance at the physical level. The basis for performing topology trimming is that removing low-energy-density elements (i.e., dead zone material or follower material) will not change the main mechanical properties of the structure, which is in line with the application logic of Saint-Venant's principle in structural simplification. In this way, this step can extract the mechanical essence structure that truly bears the rebound resistance from the complex geometric shape of the spring, which solves the technical problem of traditional methods that rely only on macroscopic geometric dimensions to calculate stiffness and cannot characterize the local stiffness characteristics of complex irregular springs in specific force directions. This ensures that the equivalent stiffness coefficient calculated subsequently has a very high physical confidence.

[0067] Step S2013: Map the bending deformation process of the main stress transmission skeleton along the motion trajectory into a three-dimensional stiffness potential energy surface, the dimensions of which include contact point displacement, lattice distortion degree, and accumulated elastic potential energy.

[0068] Using real-time ambient temperature data as a slicing plane, the stiffness potential energy surface is dimensionally reduced and cut to obtain force and displacement characteristic curves.

[0069] like Figure 3 The diagram shown illustrates the principle of dimensionality reduction and mechanical feature extraction of a three-dimensional stiffness potential energy surface.

[0070] Reference Figure 3 This illustrates the process of extracting mechanical properties at a specific temperature from a complex three-dimensional potential energy space. The figure contains three coordinate axes: The axis represents the contact displacement (horizontally to the right), which is the independent variable. It simulates the process of the moving iron core pushing the spring from 0 to 1.5mm, equivalent to the distance traveled. The axis represents the degree of lattice distortion (towards the lower left). This is a parameter that characterizes the initial deformation of the microstructure at different temperatures (e.g., -40℃, 85℃), equivalent to the environmental background. The axis represents the accumulated elastic potential energy (vertically upward), which is the dependent variable and represents the total energy stored in the system. A higher value indicates more severe deformation. The gray surface in the middle of the figure is the stiffness potential energy surface. Through the continuous spatial surface fitted by multiple sets of offline simulations, it fully records the energy response characteristics of the relay in the entire temperature range and the entire stroke, which is equivalent to a pre-mapped energy topography map. Figure 3 There is a black arrow pointing to the left. A point on the axis, labeled as This represents the real-time ambient temperature. Through function The calculated target lattice distortion value, where Let the ambient temperature be the variable. Draw a line perpendicular to this point. The plane of the axis (the slice plane shown by the dashed box in the figure) intersects the potential energy surface, and a black solid curve is extracted from it, i.e. The curve labeled is the potential energy curve at the current temperature. It is the intersection of the slice plane and the gray surface, which describes in reduced dimension the single relationship between potential energy and displacement at the current specific temperature.

[0071] Figure 3 The right side shows the transformation from the potential energy curve to the mechanical curve. This is achieved by analyzing the truncated potential energy curve. Perform geometric differentiation, i.e., find the derivative. The curve in the right coordinate system was obtained. That is, the force-displacement characteristic curve.

[0072] Reference Figure 3 Construct a parameterized feature space coordinate system independent of the geometric coordinate system. ,in The axis represents the contact displacement dimension, which serves as the independent variable to simulate the process of the moving iron core pushing the spring. Its value originates from the forced displacement boundary conditions applied by a specific end node of the principal stress transfer skeleton extracted by the finite element solver at the free end (i.e., the moving contact position). When constructing the surface, the simulator controls this end node to gradually move along the Z-axis (the direction of the moving iron core's movement) from the initial position 0 to the maximum overtravel position, for example, 1.5mm. During this process, although the displacement vectors of other nodes in the mesh... It changes accordingly, but It is the preset step value used as the independent variable. The axis represents the thermally induced lattice distortion degree. This dimension, as a parameter, characterizes the initial deformation state of the material's microstructure caused solely by temperature. This state is the environmental prerequisite before mechanical bending occurs. For different discrete temperature points, such as -40℃, 20℃, and 85℃, the thermally induced drift vectors of all nodes on the framework are calculated. , The root mean square value of the magnitudes of these thermally induced drift vectors, i.e. N is the total number of nodes. The value is a single-valued function of temperature; when constructing the surface, each fixed value... The value represents a specific thermally prestressed mesh state. The axis represents the cumulative elastic potential energy dimension, which serves as the dependent variable, and its value represents the system's energy response to mechanical deformation. For each given temperature state... (Mesh shape determined by S2011) and each displacement step size (Boundary conditions are set by S2013), using the stiffness matrix and the currently solved total field nodal displacements Calculate the strain energy of each element. Summing the strain energy of all elements within the frame. As the contact point shifts... As it grows larger, the skeleton becomes more curved. The value increases, therefore the total potential energy It increases non-linearly and cumulatively.

[0073] Through multiple sets of offline simulation iterations, for example, first fix... Scan once From 0 to 1.5, we get a line. Curve; then fix Repeated scan, among which This represents the lattice distortion value calculated when the relay is in an environment of -40°C. The lattice distortion value calculated when the relay is in an environment of 20°C will be used to represent this series of values. Discrete points are fitted to a continuous spatial surface ,Right now Figure 3 The gray surface in the image needs to have a correlation function constructed between temperature and the degree of lattice distortion. The construction method is as follows: Beforehand, a relay reed model is subjected to pure thermal stress simulation at multiple discrete temperature points ranging from -40℃ to 125℃ under no mechanical load. The lattice distortion values ​​of the basic framework caused by thermal expansion at each temperature point are extracted. The least squares method is then used to fit these discrete points to a continuous function. When running online, based on real-time temperature Using pre-built association functions Calculate the current target thermal distortion value ,like Figure 3 As shown on the left; during the real-time compensation process, the real-time ambient temperature is substituted into this function to calculate the specific lattice distortion value corresponding to the current temperature. In three-dimensional space, make a perpendicular to The axis and the intercept are The plane, i.e. Figure 3 The slice plane in the diagram intersects the three-dimensional stiffness potential energy surface to form a two-dimensional intercept, namely... Figure 3 The black solid line in This cross-section characterizes the accumulated elastic potential energy under the current temperature prestressed state. With contact displacement Change relationship Based on the principle of virtual work, the force in a conservative force field is equal to the negative gradient of the potential energy with respect to the displacement. Therefore, for this intercept function... Regarding displacement Performing first-order differentiation is... ,like Figure 3 The geometric differential process indicated by the arrow This represents the instantaneous elastic restoring force generated by the moving reed under the current thermomechanical state. and They represent the cumulative elastic potential energy. Regarding contact displacement The partial differential increment can transform the energy-displacement relationship into a force-displacement relationship, outputting the force-displacement characteristic curve at that temperature. ,like Figure 3 As shown in the coordinate system on the right.

[0074] Step S2013 maps the complex bending process of the moving spring into a three-dimensional stiffness potential energy surface and performs a dimension reduction solution, which originates from the principle of virtual work and the potential energy gradient theory in classical mechanics. In physics, the force state of an elastic body in a conservative force field has a strict differential correspondence with its stored strain energy, that is, the generalized force is equal to the partial derivative of the potential energy function with respect to the generalized displacement. The degree of lattice distortion is used as the third dimension, based on the initial strain theory in thermoelasticity. That is, although temperature changes do not directly produce mechanical displacement, they will change the energy level benchmark (thermal prestress state) inside the material, thereby changing the shape of the force-displacement curve. By constructing a continuous surface that includes displacement, thermal distortion, and potential energy, an energy state space for all working conditions is essentially established at the mathematical level. The highly nonlinear thermo-mechanical coupled finite element method (FEA) solution process has been successfully transformed into a low-computational-power lookup-slicing-differentiation geometric operation. This allows the microcontroller to cut out the most realistic mechanical property curve based on the real-time temperature within milliseconds. This avoids the computational bottleneck of online FEA simulation while retaining the accurate characterization of the influence of thermal prestress on stiffness nonlinearity, enabling real-time and accurate prediction of stiffness characteristics over a wide temperature range.

[0075] Step S2014: Perform geometric differentiation on the force and displacement characteristic curve, calculate the tangent slope of the curve near the preset operating point, and obtain the equivalent stiffness flux.

[0076] The equivalent stiffness flux is scalarized, and the equivalent stiffness coefficient is output.

[0077] The geometric differential processing refers to using numerical differential methods, such as the central difference method, to process the force-displacement characteristic curves. Slope calculation is performed to obtain the stiffness function. , u represents the geometric differential. The preset operating point refers to the displacement coordinate corresponding to the moment the relay contacts just make contact. Considering manufacturing tolerances and micro-roughness, a single contact point is insufficient to represent the overall characteristics; therefore, a definition is made based on... The small displacement neighborhood centered on , Represents the displacement coordinates at the instant the moving and stationary contacts of the relay just make contact. The radius of the small displacement tolerance centered at the point, for example, 0.05 mm, defines a local integral neighborhood covering the area before and after contact. Within this neighborhood, the stiffness function... The system presents a series of varying numerical values, and the set of these values ​​constitutes the equivalent stiffness flux, which vividly describes the fluctuating flow rate of stiffness at the instant of contact. The scalarization process is used to extract engineering-usable feature values. Specifically, it involves calculating the equivalent stiffness flux within the neighborhood by integral averaging, i.e., solving the formula... This integration operation smooths out local numerical fluctuations, and the output... The equivalent stiffness coefficient precisely quantifies the average elastic capacity of the moving spring to resist the impact of contact closure under the current specific temperature and thermal deformation skeleton state, providing accurate physical parameters for predicting subsequent rebound energy.

[0078] Step S2014 derives the equivalent stiffness coefficient through geometric differentiation and integral averaging, drawing upon numerical differentiation in mathematical analysis and the integral mean value theorem in statistics. Physically, stiffness is defined as the rate of change of force with respect to displacement; therefore, performing geometric differentiation on the characteristic curve is the only mathematical approach to obtaining instantaneous stiffness. Considering that microscopic physical contact is not an ideal single-point abrupt change but a continuous process with minute deformation zones, the tangent slope at a single geometric point is easily distorted by numerical noise, i.e., a stiffness singularity. Therefore, this step, based on the idea of ​​region averaging, introduces an equivalent stiffness flux and performs neighborhood integration. Through integral operations, stiffness fluctuations near the contact point are smoothed and filtered, removing high-frequency noise caused by mesh discretization or numerical calculations, and extracting steady-state stiffness characteristic values ​​representative of engineering applications; this coefficient... It can more realistically and robustly reflect the macroscopic elasticity of the moving spring against impact at the moment of contact, thereby ensuring that the calculation of rebound energy in subsequent steps is based on stable and reliable physical parameters, effectively preventing control misjudgment caused by parameter fluctuations.

[0079] Step S202: Based on the basic driving energy parameters and the load voltage data, construct the electromagnetic conversion efficiency function and calculate the theoretical impact energy;

[0080] Calculate the basic pulse width time Total input electrical energy of the inner coil Construct the electromagnetic conversion efficiency function The process is based on the magnetic flux density-magnetic field strength curve provided by the relay core material supplier. Due to the magnetic saturation phenomenon of magnetic materials, the excitation voltage... and current As the magnetic permeability increases, the proportion of electrical energy converted into effective magnetic field energy decreases. Through finite element electromagnetic simulation, energy conversion ratio curves under different excitation intensities are fitted, which are... Its physical meaning is the percentage of input electrical energy that is actually converted into mechanical energy to drive the moving iron core to do work. In the calculation, the real-time voltage... and the coil current estimated based on Ohm's law Substituting into the electromagnetic conversion efficiency function, we obtain the current conversion efficiency value. The formula for calculating the theoretical impact energy is: ,in The work done by the moving iron core to overcome the potential energy of the reaction spring and mechanical friction during its entire stroke is a parameter in existing technology. This value is obtained by plotting the force-displacement static characteristic curve of the relay using a force gauge in a standard laboratory environment, and then integrating the area under the curve. This area is then stored as a constant or displacement function. The final result... This refers to predicting the remaining kinetic energy of the moving iron core at the moment of closed contact, which may lead to a rebound.

[0081] Step S203: Substitute the equivalent stiffness coefficient and theoretical impact energy into the collision recovery coefficient model to calculate the expected rebound potential energy after the collision. If the expected rebound potential energy after the collision is greater than the minimum attraction barrier, the current state is marked as a high-risk rebound zone, and the critical rebound energy threshold is output.

[0082] The collision restitution coefficient model includes a stiffness-restitution coefficient mapping formula and a kinetic energy recovery equation formula, which are analyzed in detail below: In order to eliminate the difference in physical dimensions and establish a relative reference frame for stiffness, preset stiffness boundary parameters are called. and These two parameters represent the theoretical stiffness extremities of the relay reed at preset extreme high and low temperatures, respectively. The preset extreme high temperature, such as 125°C, results in the lowest material modulus, while the extreme low temperature, such as -40°C, results in the highest material modulus. The equivalent stiffness coefficient output in step S2014 is then calculated using a maximum-minimum normalization algorithm. The processing and calculation formula is as follows: Output dimensionless stiffness exponent The value ranges from 0 to 1; substitute this index into the preset stiffness-restitution coefficient mapping formula. ,in Stiffness-restore factor The baseline rebound coefficient of the material is obtained by heating a sample of the contact material to a preset maximum operating temperature in a laboratory environment, bringing it to its softest physical state. A miniature Hopkinson bar or a high-speed camera is used to record the rebound velocity and incident velocity under a standard impact. The ratio of these two values ​​is then calibrated. ;and and The dimensionless material constant is obtained through multiple experiments and fitting at different temperature points, and is used to describe the change in stiffness exponent. The increase is due to the physical laws governing the reduction in plastic dissipation and the nonlinear increase in the rebound coefficient caused by material hardening; this is achieved using the kinetic energy recovery equation formula. The formula for calculating the expected rebound potential energy after a collision is based on the kinematic definition of the coefficient of restitution. To divide the rebound velocity into the incident velocity, and since kinetic energy is proportional to the square of the velocity, the energy retained after the rebound is... equal to incident energy Multiply by the square of the coefficient. The minimum attraction barrier. The method for obtaining the static electromagnetic attraction curve of the relay in the closed state is as follows: through finite element electromagnetic simulation or force gauge experiment, the static electromagnetic attraction curve of the relay in the closed state is obtained. With the characteristic curve of the reaction spring Integrate the difference between the two over the contact opening distance, i.e. The integral variable Refers to the displacement coordinate along the Z-axis defined in the coordinate system of step S2011. and These represent the Z-axis coordinates of the closed dead position and the physically open position, respectively. The physical meaning of this integral value is the minimum mechanical energy required to forcibly pull the moving iron core from the closed position to the open position along the Z-axis, i.e., the magnetomotive force trap depth, in order to overcome the work done by the spring. The system executes a logic comparison; if... This means that the rebound kinetic energy is sufficient to overcome the magnetic attraction force, causing the contacts to separate instantaneously. The system determines this as a high-risk rebound and, based on the critical equilibrium condition... Reverse calculation, output critical bounce energy threshold This serves as the constraint boundary for subsequent control algorithms.

[0083] Step S203 quantifies the energy dissipation and conversion law of the moving contact at the moment of closure by using the collision recovery coefficient model, and uses the stiffness coefficient to nonlinearly correct the recovery coefficient, thereby accurately predicting the magnitude of the rebound potential energy. Combined with the minimum attraction barrier obtained from finite element simulation, i.e. the minimum mechanical work required for the moving iron core to detach from the magnetic attraction, an energy criterion for rebound is established. This step is indispensable in the overall method because it not only realizes the leap from static stiffness to dynamic rebound risk, but also provides a clear quantitative upper limit for subsequent drive energy control by setting a critical rebound energy threshold. It directly determines whether the system needs to start derating drive or damping suppression strategy, and is the decision center for preventing secondary rebound disconnection failure. It ensures that the system can actively identify and avoid high-risk operating conditions in a wide temperature range, thereby greatly improving the engagement reliability of the relay in extreme environments.

[0084] The method for dynamic compensation of magnetic properties of magnetic latching relays adapted to wide temperature range environments also includes the following steps:

[0085] Step S300: Based on the critical rebound energy threshold, construct an ideal convergence trajectory that satisfies the physical deceleration constraint, and perform inverse dynamics solution to generate a composite step drive waveform command;

[0086] The step S300 step of generating the composite step drive waveform instruction specifically includes the following steps:

[0087] Step S301: Construct the displacement-velocity phase plane of the moving iron core. Using the time-domain parameterized mapping method, combined with the initial disconnection state, the contact point just contacting state, and the soft landing velocity condition, construct the motion equations of the moving iron core and generate the ideal convergence trajectory.

[0088] Construct the real-time displacement of the moving iron core along the Z-axis. The instantaneous speed of the horizontal axis and the moving iron core The vertical axis is a two-dimensional rectangular coordinate system, i.e., the displacement-velocity phase plane, and the Z-axis direction is the suction motion direction defined in step S2011; obtain the ideal convergence trajectory. The specific method is the time-domain parameterized mapping method, which constructs a cubic polynomial function of the displacement of the moving iron core as a function of time. , to These are the four coefficients of a polynomial to be determined. To uniquely determine the four coefficients of this polynomial, four explicit physical boundary conditions are set: First, the initial displacement condition, when... When, displacement , The first is the initial disconnection position coordinate, which is the absolute position of the moving iron core on the Z-axis when the relay is in a static released state under the action of the return spring, corresponding to the zero point of the motion time; the second is the initial speed condition, when... At that time, speed First, a stationary start; second, the termination displacement condition, when... When, displacement , The coordinates represent the position of the moving iron core on the Z-axis at the instant the moving and stationary contacts make physical contact after eliminating the air gap, corresponding to the end of the ideal engagement time. ,in The preset ideal engagement time, for example, 5ms; fourth, the soft landing speed condition, when At that time, speed , The preset maximum allowable contact speed threshold is where The value is determined by the critical rebound energy threshold. Combining the quality of the moving iron core Using the kinetic energy formula Determine the initial solution and After establishing the relationship, perform a physical deceleration capability verification: calculate the instantaneous acceleration function corresponding to the trajectory. , where the symbol This represents the second derivative of the displacement function with respect to time. Its physical meaning is the instantaneous acceleration value required by the moving iron core at any moment along the planned path, and it also provides the maximum reverse braking acceleration of the electromagnetic mechanism. The method for obtaining this value is as follows: Based on Newton's second law, the maximum reverse electromagnetic attraction force that the coil can generate under rated voltage is vector-superimposed with the reaction spring resistance under the current displacement to obtain the maximum net braking force. The maximum net braking force is then divided by the mass of the moving iron core. The maximum reverse braking acceleration of the electromagnetic mechanism was calculated. It was then checked whether the condition was met throughout the entire time domain. , Represents the absolute value of the instantaneous acceleration function, if it exists. The time indicated indicates the currently set engagement time. The excessively short length caused the deceleration requirement to exceed the physical limits, so the system automatically extended it. Then, the coefficients are recalculated until the trajectory satisfies the physical constraint. After solving the motion equations of the moving iron core by simultaneously solving the above four conditions to obtain the coefficients, the time... Differentiation yields the velocity function By eliminating the time variable ,Will and In this correspondence, a monotonically smooth function curve is generated on the phase plane. The physical significance of this trajectory lies in the fact that it represents the maximum kinetic energy boundary that the moving iron core can possess at each displacement point, under the premise of balancing rapid engagement and impact-free soft landing.

[0089] Step S302: Discretize the motion stroke of the moving iron core, and determine the target kinetic energy increment within each discrete displacement step based on the ideal convergence trajectory;

[0090] Based on the target kinetic energy increment, a work-energy balance equation is constructed, the net electromagnetic force is calculated, and the duty cycle control sequence of the driving pulse is obtained by combining the current load voltage and the real-time coil resistance.

[0091] Discretize the entire motion path, for any current position in the path and the next position ,in This is the step size for small displacements, specifically set for numerical discretization. The index number is used to determine the current ideal velocity based on the ideal trajectory. and the speed of the next target ,in This represents the ideal velocity at the current position. Indicates the ideal speed for the next position; through the mass parameters of the moving iron core Calculate the target kinetic energy increment within this step size. The work done by the net external force on the moving iron core during its motion is equal to the increase in kinetic energy. This net external force includes the driving force (electromagnetic force) and the resistance force (reaction force, spring force, and friction). Therefore, the work-energy balance equation is established as follows: ,in This is the resistance force of the reaction spring at the current displacement, a value obtained from the pre-stored reaction spring characteristic curve. The average frictional force is determined by the work done by friction. Divide by the total distance to obtain the net electromagnetic force required; calculate the net electromagnetic force required. ; Referencing the analytical formula for electromagnetic attraction ,in The electromagnetic attraction generated when a coil is energized. For real-time load voltage, The real-time coil resistance is a function constructed based on the real-time temperature, where... Let L be the rate of change of inductance with displacement z, an inherent parameter of the system, and D be the duty cycle of the driving pulse to be solved, used to adjust the equivalent driving voltage of the coil; let , will the known Substituting the relevant parameters into the formula, the corresponding drive pulse duty cycle is solved in reverse. It iterates through all discrete points in sequence to generate a duty cycle control sequence that precisely matches the drive pulse of the current operating condition. .

[0092] Step S303: Perform time-domain discretization processing on the duty cycle control sequence to generate the composite step drive waveform instruction, which includes a high-energy acceleration in the front section, inertial coasting in the middle section, and PWM chopping speed limiting in the final section.

[0093] The duty cycle control sequence varies with displacement. Mapping instructions to time-varying waveforms The mapping method is to use kinematic formulas. By accumulating time point by point, the execution time and duration of each duty cycle value on the time axis are determined. Based on this, according to the dynamic characteristics of the moving iron core at different stroke stages, the full-time domain command is divided into three logical stages: the first stage is "high-energy acceleration in the early stage," and its determination is based on... From the start-up moment until the moving iron core speed reaches the peak of the ideal trajectory Within this interval, the peak of the ideal trajectory is the ideal convergent trajectory. At the maximum speed, during this stage, the system forcibly overrides the output. The purpose of the full-voltage pulse is to generate a strong magnetic field instantaneously using the maximum ampere-turns of the coil to overcome static friction and the initial preload of the reaction spring, thus eliminating the starting delay; the second stage is "mid-stage inertial coasting," which is determined by the moving iron core passing through the acceleration phase and displaced. In the total journey In the interval where the moving iron core crosses the acceleration phase, its instantaneous velocity first reaches the global maximum velocity planned by the ideal convergence trajectory. At this time, the compression of the reaction spring is still small, but the moving iron core has gained a large amount of kinetic energy. During this stage, the system executes a low-energy-level maintenance strategy, increasing the duty cycle... Reduce to the minimum magnetomotive force required to maintain motion, such as a duty cycle control sequence that varies with displacement. The lower value is usually in In the middle stage, the inertial momentum of the moving iron core is used to overcome the spring resistance and slide, preventing excessive accumulation of unnecessary kinetic energy in the middle of the stroke; the third stage is "final stage PWM chopping speed limiting", which is determined by displacement. Entering the distance from the closed contact position at last In the sensitive region, at this stage, the system strictly executes the high-precision duty cycle sequence derived from the inverse solution. By rapidly chopping the coil voltage using high-frequency PWM, electronic braking is essentially implemented. This involves actively reducing the duty cycle to weaken the net attraction force as the air gap decreases and the electromagnetic attraction force naturally increases non-linearly. This ensures that the acceleration generated by the net external force is strictly controlled. The net external force is composed of attraction force, elastic force, and resistance, thereby forcing the moving iron core to conform to... Soft contact is achieved at minute speeds, thereby generating a sequence containing these three feature segments. The sequence is the composite step drive waveform instruction.

[0094] Step S303 maps the duty cycle sequence based on displacement planning onto the time axis and constructs a phased composite drive waveform based on the physical processes of the moving iron core's initial acceleration, mid-course coasting, and final soft landing. The initial full-voltage output aims to quickly establish magnetic flux to overcome static friction, the mid-course low-energy maintenance utilizes inertial coasting to reduce energy consumption, and the final PWM chopping weakens electromagnetic attraction to achieve speed clamping. This step is indispensable in the overall method because it transforms the theoretical energy control strategy into specific electrical signal instructions that the microcontroller can execute, directly determining the voltage application method across the coil. This not only ensures that the moving iron core can move strictly along the ideal convergence trajectory to achieve impact-free engagement, but also effectively reduces coil temperature rise and contact wear through refined energy management. It is a key execution link for achieving the soft landing goal and extending the life of electrical appliances, ensuring the accurate implementation of the control algorithm at the physical level.

[0095] The method for dynamic compensation of magnetic properties of magnetic latching relays adapted to wide temperature range environments also includes the following steps:

[0096] Step S400: Execute the composite step drive waveform command to drive the relay coil, collect the coil's terminal voltage and current response in the drive gap window, and calculate the motion state vector of the moving iron core in real time through the flux linkage observer;

[0097] The microcontroller's timer module is based on time-domain waveform instructions. The MOSFETs in the H-bridge drive circuit are switched to apply corresponding voltage pulses to the relay coil. During this process, to avoid interference from PWM switching noise on the analog signal sampling, a "drive gap window" is set as the data acquisition time. This window is defined as the level-stable region within the PWM carrier cycle, such as the center alignment moment of the PWM waveform or the middle moment of the freewheeling phase. Within this window, a high-precision ADC module is used to synchronously acquire the real-time terminal voltage across the coil. and the real-time excitation current flowing through the coil These two physical quantities constitute the input vector of the observer, reflecting the real-time electrical response of the electromagnetic system under the current driving command.

[0098] Step S400, which involves real-time calculation of the motion state vector of the moving iron core using a flux linkage observer, specifically includes the following sub-steps:

[0099] Step S401: Establish the dynamic voltage balance equation of the relay, and use the collected terminal voltage and current response to separate the coil resistance voltage drop and inductance voltage drop, and extract the induced back electromotive force component.

[0100] An electrical mathematical model of a relay coil circuit is constructed based on Kirchhoff's voltage law, and its time-domain expression is as follows: ;in, The determined coil resistance including temperature compensation corresponds to step S100; For nonlinear incremental inductance, this parameter is obtained by scanning different current intensities through finite element electromagnetic simulation during the offline calibration phase. and different air gap positions The coil flux linkage is used to establish a three-dimensional inductance lookup table, which is stored in the controller. During real-time calculations, the real-time current is used... And the displacement of the moving iron core estimated at the previous moment According to the table, the so-called stripping process is essentially a numerical subtraction operation, and the system calculates the voltage drop component. With inductive voltage drop component ,in The derivative of the current with respect to time is given, and these two components are derived from the total terminal voltage acquired in real time. Subtracting from the middle, we can extract the component of the back electromotive force induced by motion, which is generated solely by the moving iron core cutting magnetic field lines. This component is the only physical signal source for subsequent calculations of the invisible mechanical motion state.

[0101] Step S402: The extended Kalman filter algorithm is used to suppress noise and estimate the state of the induced back electromotive force component. The corrected coil resistance value is obtained by estimating the coil resistance drift error in real time.

[0102] Construct a nonlinear state-space model in the discrete-time domain and define the state vector. ,in represent The actual state variable of the coil current at any given time. represent The real state variable of the flux linkage at any given time. represent The resistance thermal drift error term to be estimated at any given time; setting the observation equation. ,in This refers to the data actually acquired by the ADC module within the drive gap window in step S400. The physical measurement value of the coil current at time step; the extended Kalman filter algorithm periodically performs prediction-correction iterations, specifically in the prediction phase, using the posterior state estimate from the previous time step. and the terminal voltage input acquired at the current moment , It includes the optimal current estimate output by the algorithm at the previous time step. Optimal flux linkage estimation and resistance error and the terminal voltage input acquired at the current moment To deduce the prior state estimate at the current moment. The specific current prediction formula is as follows: In the formula For the present The prior estimate of the coil current at time . The subscript represents the motion-induced back electromotive force component calculated based on the model. Indicates the current discrete sampling time, subscript This indicates the time of the previous discrete sampling. The discrete sampling period for the microcontroller to execute the algorithm, i.e., the time interval between two adjacent calculations, for example... L represents the nonlinear incremental inductance; during the correction phase, the innovation residual is calculated. This residual quantifies the difference between the actual measured current and the model predicted current. The deviation between them; using the Kalman gain matrix The prior state vector is weighted and corrected using the following formula: This yields the final posterior state estimate at the current moment. Through this closed-loop analysis, the prediction deviation is continuously corrected using the actual measured current, and the posterior estimate of the resistance error at the current moment is output in real time. Output corrected resistance value This value eliminates the effects of temperature drift and is passed as a high-precision physical parameter to subsequent steps.

[0103] Step S403: Integrate the voltage equation based on the corrected coil resistance value to obtain the real-time flux linkage value. Calculate the real-time air gap distance and instantaneous velocity of the moving iron core relative to the stationary contact based on the real-time flux linkage value, and combine them to generate the motion state vector.

[0104] The flux linkage is calculated using Faraday's law of electromagnetic induction, based on the corrected coil resistance value. The voltage equation across the coil is integrated in the time domain, and the calculation formula is as follows: The real-time flux linkage value obtained by integration This represents the true magnetic flux level of the current physical system. This indicates the real-time acquired terminal voltage input. This represents the real-time acquired current. The system calls a pre-stored three-dimensional data mapping table of "magnetic flux linkage-current-displacement". It should be noted that this mapping table is pre-constructed based on the inherent electromagnetic characteristics of the relay and belongs to the prior art in this field. The acquisition method is usually as follows: in the offline stage, a three-dimensional solid model of the relay is established using finite element electromagnetic simulation software, and a full-domain parametric scan of the coil current and moving iron core displacement is performed to calculate the corresponding static magnetic flux linkage value at each "current-displacement" combination point, thereby generating this three-dimensional lookup table. This is a common technique in the field of electromagnetic actuator modeling and will not be elaborated here. In real-time operation, the real-time acquired current... The high-precision magnetic flux obtained by the above integration The actual displacement of the moving iron core is calculated in reverse using a bilinear interpolation algorithm, with the input being a bivariate index to the mapping table. ; Using the formula for velocity calculation based on back electromotive force ,in To induce the back electromotive force component, the denominator term... This represents the partial derivative of the magnetic flux linkage with respect to displacement. This coefficient is obtained based on the actual displacement just calculated. and real-time magnet links The current working point is determined by performing tangent slope calculation or difference operation on the data at that working point in the three-dimensional data mapping table; the calculated actual displacement is then... Compared with actual speed Combined and encapsulated into a two-dimensional column vector This is the motion state vector, which will be used as the real-time monitoring object input to the subsequent dynamic magnetic potential trap constraint controller. It is specifically used to identify whether the moving iron core has mechanically rebounded and the magnitude of the rebound kinetic energy, thereby determining whether to trigger the active magnetic damping suppression mechanism.

[0105] Step S403 obtains a high-confidence real-time flux linkage by performing time-domain integration on the corrected voltage equation. Using flux linkage and current as a dual-variable index, the displacement of the moving iron core is solved by looking up a table. The instantaneous velocity is calculated based on the relationship between the back electromotive force and the partial derivative of the flux linkage. Thus, the true motion state of the moving iron core inside the invisible magnetic field can be reconstructed without the need for external sensors. This step is indispensable in the overall method because it provides the real-time feedback signal required for closed-loop control, enabling the system to "see" the microscopic motion trajectory of the moving iron core. It is particularly important for capturing the tiny displacement fluctuations and velocity changes at the moment of closure, providing a unique and accurate physical criterion for the subsequent triggering of active magnetic damping. This breaks the limitation of traditional open-loop control in being unable to perceive the rebound phenomenon and is the perception basis for realizing intelligent closed-loop correction.

[0106] The method for dynamic compensation of magnetic properties of magnetic latching relays adapted to wide temperature range environments also includes the following steps:

[0107] Step S500: Based on the motion state vector, mechanical rebound is detected and active magnetic damping adjustment is triggered to forcibly clamp the moving iron core within the magnetically maintained balance region;

[0108] Furthermore, the analytical steps for identifying mechanical rebound and triggering active magnetic damping adjustment are as follows:

[0109] Step S501: Monitor the velocity component in the motion state vector in real time. When the velocity sign changes from positive to negative and the displacement is within the closed interval of the contact point, it is determined that a mechanical rebound has occurred, and the rebound acceleration is calculated.

[0110] The motion state vector A logic discrimination unit is introduced, and the contact closure interval is set as follows: ,in To account for minute displacement tolerances in mechanical tolerances, such as 0.05 mm, this range covers the process position after the moving and stationary contacts make physical contact; the system scans the actual velocity components at an extremely high sampling frequency (e.g., 20 kHz). During the normal sigma activation process, symbolic logic comparison is performed. It should always be positive or zero, with positive values ​​pointing in the closed direction; it is only true if both of the following conditions are met simultaneously: firstly, the displacement components... Within the contact closure range, the second is the velocity component. The sign of the arithmetic symbol changes abruptly from positive to negative. The system determined that the moving iron core had experienced a mechanical rebound, meaning that the moving iron core had detached from the magnetically held position due to the rigid impact test. The function is a sign function; the system calls the differential operation module, selects several adjacent sampling points before and after the velocity reversal moment (e.g., 3 points), and uses the central difference method to calculate the acceleration value at the instant of rebound. , The historical velocity value of the moving iron core recorded and stored in the register at the previous sampling moment is used as a calculation benchmark for comparison with the current velocity; In mathematics, it represents the derivative of velocity with respect to time, and in physics, it is the instantaneous acceleration. In digital control systems, it is discretized and solved using the above-mentioned difference quotient. This acceleration value objectively quantifies the intensity of the rebound burst force and provides a mechanical basis for subsequent calculation of damping force.

[0111] Step S502: Based on the rebound acceleration and the mass of the moving iron core, calculate the reverse Lorentz force required to counteract the rebound momentum, and calculate the required damping current amplitude according to the current magnetic circuit saturation.

[0112] A rebound suppression dynamics model is constructed based on Newton's second law to calculate the target mechanical braking force required to suppress the rebound trend in a very short time. ,in A preset safety redundancy factor, such as 1.2, is used to ensure that the generated magnetic attraction force is sufficient to completely cover the mechanical rebound force. To determine the coil current required to generate this force, the system must solve the problem of "magnetic saturation nonlinearity": when rebound occurs, the moving iron core is in a closed position, the magnetic circuit reluctance is extremely low, and a large current needs to be injected to generate braking force. At this time, the iron core material is very prone to entering the magnetic saturation region, causing the electromagnetic force and current to no longer obey a simple square relationship. Therefore, the current magnetic circuit saturation is not obtained through additional sensors, but directly from the real-time flux linkage value obtained by integration. To characterize this, the magnetic flux linkage is the integral of the magnetic induction intensity, directly reflecting the degree of deflection of the magnetic domains inside the iron core. The system calls the preset "force-current-magnetic saturation coefficient mapping table," which is constructed based on the nonlinear constitutive relationship of ferromagnetic materials. The method for obtaining this table is as follows: in the offline stage, using finite element electromagnetic simulation, the moving iron core is fixed in the closed position, and the excitation current of the coil is scanned across the entire domain from zero to overload, recording each discrete scan excitation current point. The corresponding static flux linkage value and the generated scanning electromagnetic attraction This generates a target electromagnetic force. and current flux linkage state Input index, with the required drive current This is the output three-dimensional lookup table; in real-time control, the controller will calculate the target inverse Lorentz force. (as) (Input) and current real-time magnet links Substituting into the table, bilinear interpolation can be used to accurately find the damping current amplitude that can generate sufficient braking force at the current saturation level. .

[0113] Step S503: Control the H-bridge drive circuit to flip the voltage polarity within a preset microsecond time window, pump the coil current to the damping current amplitude, force the moving iron core to clamp in the magnetic holding balance area, and complete the closed-loop correction of the attraction action.

[0114] Based on the calculated damping current This immediately triggers the highest priority active magnetoresistive interrupt, opening a preset microsecond-level time window, typically set to... to Within this window, the controller sends a command to the H-bridge drive circuit to reverse the polarity of the voltage applied across the coil, causing it to generate a strong electromagnetic attraction pointing towards the dead center. This direction is mechanically strictly opposite to the direction in which the moving iron core attempts to bounce outward. The system rapidly pumps the coil current to [a specific value] through high-frequency PWM modulation. This ensures that before the moving iron core produces a significant rebound displacement, i.e., before the actual rebound distance is less than the preset safety threshold, such as... Previously, this enormous instantaneous electromagnetic force was used as a magnetic clamping force to counteract the reverse momentum generated by the mechanical collision, forcibly displacing the moving iron core. Locked in Within the nearby magnetically balanced region. The motion state vector of the moving iron core is monitored in real time; when the instantaneous velocity is simultaneously satisfied... and displacement fluctuation variance When these two conditions are met, it is determined that the displacement fluctuation has converged. The preset threshold for determining stillness is, for example... , A preset steady-state convergence threshold is used to determine whether the moving iron core has stopped oscillating. Based on the convergence of displacement fluctuations, to avoid sudden current cutoff causing new electromagnetic shocks, the system executes a gradual withdrawal strategy, i.e., controlling the PWM duty cycle to smoothly decrease with a preset linear slope, such as every... reduce Until the current returns to zero, the relay seamlessly and smoothly transitions to the static holding state of the permanent magnet, thus completely eliminating the hidden danger of "secondary rebound disconnection" at the physical level.

[0115] Step S503 rapidly reverses the coil voltage polarity within a microsecond-level time window when mechanical rebound occurs, generating a strong electromagnetic attraction force strictly opposite to the direction of the rebound momentum. This clamping force counteracts the tendency of the moving iron core to detach, and after the motion converges, a smooth exit is achieved through linearly decaying current, thereby physically forcing the moving iron core to stabilize in the magnetic holding region. This step is indispensable in the overall method, serving as the last line of defense against secondary rebound disconnection. It acts directly at the moment the fault occurs, actively intervening to eliminate the violent rebound that cannot be suppressed by passive attraction alone. It completely solves the problem of engagement failure caused by stiffness changes in a wide temperature range, ensuring that the relay can still achieve reliable electrical connection under extreme operating conditions. It is the core correction method to ensure the final execution performance of the system.

[0116] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0117] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0118] The foregoing description is illustrative of the invention and should not be construed as limiting it. Although several exemplary embodiments of the invention have been described, those skilled in the art will readily understand that many modifications can be made to the exemplary embodiments without departing from the novel teachings and advantages of the invention. Therefore, all such modifications are intended to be included within the scope of the invention as defined in the claims. It should be understood that the foregoing description is illustrative of the invention and should not be construed as limiting it to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The invention is defined by the claims and their equivalents.

Claims

1. A method for dynamic compensation of magnetic properties of a magnetic latching relay adapted to a wide temperature range environment, characterized in that, Includes the following steps: Acquire real-time temperature data of the current working environment of the relay and real-time load voltage data of the power supply, construct an initial magnetic circuit drive model based on temperature-voltage dual variables, and output basic drive energy parameters. Based on the real-time temperature data, a three-dimensional stiffness potential energy surface is constructed to calculate the equivalent stiffness coefficient. Combined with the basic driving energy parameters, the mechanical collision kinetic energy at the moment of closing of the moving contact is predicted, and the critical rebound energy threshold is generated. Based on the critical rebound energy threshold, an ideal convergence trajectory that satisfies the physical deceleration constraint is constructed, and inverse dynamics solution is performed to generate a composite step drive waveform command. The composite stepped drive waveform command is executed to drive the relay coil. The terminal voltage and current response of the coil are collected in the drive gap window, and the motion state vector of the moving iron core is calculated in real time through the flux linkage observer. Based on the motion state vector, mechanical rebound is detected and active magnetic damping adjustment is triggered to forcibly clamp the moving iron core within the magnetically maintained balance zone. The analytical steps for constructing the three-dimensional stiffness potential energy surface and calculating the equivalent stiffness coefficient are as follows: Based on the thermal drift vector, the thermal solid mesh is reconstructed, and the principal stress transfer skeleton is extracted through topology trimming to construct a three-dimensional stiffness potential energy surface. The equivalent stiffness coefficient is calculated by using real-time temperature to extract dimensionality reduction features from the three-dimensional stiffness potential energy surface. The analytical steps for calculating the equivalent stiffness coefficient are as follows: The relay moving spring is discretized into a lattice topological mesh, and the thermal drift vector calculated based on the real-time ambient temperature is superimposed to reconstruct the thermal solid mesh that undergoes microscopic deformation. Based on the direction of motion of the moving iron core, the stiffness contribution analysis and topology trimming are performed on the hot solid mesh to extract the main stress transmission skeleton connecting the fixed end and the contact end. The bending deformation process of the principal stress transmission skeleton along the motion trajectory is mapped as a three-dimensional stiffness potential energy surface, the dimensions of which include contact point displacement, lattice distortion degree, and accumulated elastic potential energy. Using real-time ambient temperature data as a slicing plane, the three-dimensional stiffness potential energy surface is dimensionally reduced and cut to obtain force and displacement characteristic curves. Geometric differentiation is performed on the force-displacement characteristic curve to calculate the slope of the tangent line near the preset operating point, thereby obtaining the equivalent stiffness flux. The equivalent stiffness flux is scalarized, and the equivalent stiffness coefficient is output.

2. The method for dynamic compensation of magnetic properties of a magnetic latching relay adapted to a wide temperature range environment according to claim 1, characterized in that, The analytical steps for generating the critical rebound energy threshold are as follows: Based on the basic driving energy parameters and the load voltage data, an electromagnetic conversion efficiency function is constructed to calculate the theoretical impact energy. Substitute the equivalent stiffness coefficient and theoretical impact energy into the collision recovery coefficient model to calculate the expected rebound potential energy after the collision; If the expected rebound potential energy after the collision is greater than the minimum attraction barrier, the current state is marked as a high-risk rebound zone, and the critical rebound energy threshold is output.

3. The method for dynamic compensation of magnetic properties of a magnetic latching relay adapted to a wide temperature range environment according to claim 2, characterized in that, The analytical steps for constructing an ideal convergence trajectory that satisfies the physical deceleration constraint are as follows: A displacement-velocity phase plane of the moving iron core is constructed. Using the time-domain parameterized mapping method, combined with the initial disconnection state, the contact just-contact state, and the soft landing velocity condition, a set of motion equations for the moving iron core is constructed to generate an ideal convergent trajectory.

4. The method for dynamic compensation of magnetic properties of a magnetic latching relay adapted to a wide temperature range environment according to claim 3, characterized in that, The analysis steps for performing inverse dynamics solution to generate composite step drive waveform commands are as follows: The motion stroke of the moving iron core is discretized, and the target kinetic energy increment within each discrete displacement step is determined based on the ideal convergence trajectory. Based on the target kinetic energy increment, a work-energy balance equation is constructed, the net electromagnetic force is calculated, and the duty cycle control sequence of the driving pulse is solved by combining the current load voltage and the real-time coil resistance. The duty cycle control sequence is discretized in the time domain to generate the composite step drive waveform instruction, which includes a high-energy acceleration in the front section, inertial coasting in the middle section, and PWM chopping speed limiting in the final section.

5. The method for dynamic compensation of magnetic properties of a magnetic latching relay adapted to a wide temperature range environment according to claim 1, characterized in that, The analysis steps for real-time calculation of the motion state vector of the moving iron core using a flux linkage observer are as follows: Establish the dynamic voltage balance equation of the relay, and use the collected terminal voltage and current response to separate the coil resistance voltage drop and inductance voltage drop, and extract the induced back electromotive force component. The extended Kalman filter algorithm is used to suppress noise and estimate the state of the induced back electromotive force component. The corrected coil resistance value is obtained by estimating the coil resistance drift error in real time.

6. The method for dynamic compensation of magnetic properties of a magnetic latching relay adapted to a wide temperature range environment according to claim 5, characterized in that, The analysis step of calculating the motion state vector of the moving iron core in real time through the flux linkage observer also includes: The voltage equation is integrated based on the corrected coil resistance value to obtain the real-time flux linkage value. The real-time air gap distance and instantaneous velocity of the moving iron core relative to the stationary contact are calculated based on the real-time flux linkage value, and the motion state vector is generated by combining them.

7. The method for dynamic compensation of magnetic properties of a magnetic latching relay adapted to a wide temperature range environment according to claim 1, characterized in that, The analytical steps for forcibly clamping the moving iron core within the magnetically maintained balance region are as follows: The velocity component in the motion state vector is monitored in real time. When the velocity sign changes from positive to negative and the displacement is within the closed interval of the contact point, it is determined that a mechanical rebound has occurred, and the rebound acceleration is calculated. Based on the rebound acceleration and the mass of the moving iron core, the reverse Lorentz force required to counteract the rebound momentum is calculated, and the required damping current amplitude is calculated based on the current magnetic circuit saturation. The control H-bridge drive circuit flips the voltage polarity within a preset microsecond time window, pumps the coil current to the damping current amplitude, and forces the moving iron core to clamp in the magnetic holding balance area, thus completing the closed-loop correction of the attraction action.

Images