Automobile comprehensive performance detection method and system

Abstract

The present application relates to the technical field of automobile comprehensive performance detection, and more particularly to a method and system for detecting the comprehensive performance of an automobile, which comprises obtaining a secondary magnetic field signal formed by a response excitation magnetic field of a belt layer inside a tire, and obtaining a voltage analog array representing dynamic spatial displacement of the belt layer; synchronously obtaining an acoustic emission signal of a contact interface between the tire and a roller of a dynamometer, and obtaining a time domain signal representing a transient viscoelastic state of the interface. By using tomographic inversion, time-frequency analysis and mechanical mapping, real-time radial expansion physical deformation variables and viscoelastic shear tangent stiffness values are respectively solved. Based on a dynamic flow form topological space, cross-physical field decoupling is performed, viscoelastic softening interference is stripped, and a structural expansion component is extracted. Finally, a rotational speed conversion compensation coefficient is generated, and the original rotational speed pulse signal is resampled and compensated. The present application realizes dynamic alignment of the roller linear speed and the vehicle wheel edge rotational speed, and improves the accuracy of vehicle performance evaluation under high dynamic conditions.

Description

Technical Field

[0001] This invention relates to the field of automotive comprehensive performance testing technology, and in particular to automotive comprehensive performance testing methods and systems. Background Technology

[0002] Comprehensive vehicle performance testing provides an in-depth assessment of a vehicle's power, economy, safety, and emissions levels, reflecting its overall operational status and structural integrity. The various indicators are interconnected and work together to generate comprehensive test data that provides a scientific basis for studying the maintenance and performance evolution patterns throughout the vehicle's lifecycle. In the application scenario of high dynamic response chassis dynamometer simulating high-speed cruising and extreme acceleration performance testing of vehicles, the existing high dynamic response chassis dynamometer testing technology has the following technical pain points: As the core medium for power transmission, the tire is subjected to the strong centrifugal force generated by continuous high-speed rotation, combined with the expansion of gas inside the tire caused by frictional heat and the softening of material physical properties, which causes the tire crown to generate nonlinear radial displacement outward, directly changing the effective rolling radius of the wheel during dynamic operation. Since the speed conversion logic preset by the dynamometer system is usually based on static geometric parameters, the distance from the actual rotation center of the tire to the contact surface under high-speed or high-acceleration conditions has increased significantly. This causes the linear velocity pulses collected by the dynamometer roller to be unable to accurately correspond to the actual rotation state of the drive wheel flange, resulting in inconsistencies in the timing of dynamic characteristic quantities during the sampling stage. This is reflected in the test results as a significant calculation deviation between the dynamic response characteristics and the speed change trend, seriously interfering with the objective assessment of the dynamic response characteristics of the whole vehicle under extreme operating boundaries. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a comprehensive vehicle performance testing method and system. This invention solves the technical problem of inaccurate deviation between the linear velocity of the dynamometer roller and the actual wheel-side speed of the vehicle during the acquisition of vehicle dynamic parameters due to the nonlinear radial expansion of the tire caused by the coupling effect of centrifugal force and temperature rise under high-speed rotation.

[0004] To solve the above-mentioned technical problems, the specific contents of the present invention are as follows: In a first aspect, the comprehensive performance testing method for automobiles provided by the present invention includes: Step 1: Obtain the secondary magnetic field signal generated by the excitation magnetic field in response to the belt layer inside the tire, and obtain a voltage analog quantity array characterizing the dynamic spatial displacement of the belt layer inside the tire. Step 2: Obtain the acoustic emission signal at the contact interface between the tire and the dynamometer roller to obtain a time-domain signal characterizing the transient viscoelastic state of the contact interface. Step 3: Perform tomographic inversion on the voltage analog array to obtain the real-time radial expansion physical deformation of the belt layer inside the tire; Step 4: Perform time-frequency analysis and mechanical mapping on the time-domain signal to obtain the viscoelastic shear tangent stiffness value of the rubber at the contact interface. Step 5: Time synchronization and cross-physical field decoupling are performed on the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​to obtain the structural expansion components; Step 6: Generate a speed conversion compensation coefficient based on the structural expansion component, and use the speed conversion compensation coefficient to compensate the original speed pulse signal fed back by the dynamometer roller, so as to correct the deviation between the linear velocity of the dynamometer roller and the actual wheel-side speed of the vehicle.

[0005] Furthermore, the comprehensive vehicle performance testing method of the present invention, wherein obtaining the secondary magnetic field signal formed by the excitation magnetic field in response to the internal belt layer of the tire, and obtaining a voltage analog quantity array characterizing the dynamic spatial displacement of the internal belt layer of the tire, includes: A low-frequency alternating magnetic field that penetrates the outer insulating rubber medium of the tire is emitted by a low-frequency alternating magnetic field emitting unit located inside the dynamometer drum as the excitation magnetic field to induce eddy currents on the surface of the conductive steel wire skeleton inside the tire. The secondary magnetic field signal derived from the eddy current is received by a penetrating giant magnetoresistive receiver matrix; the received secondary magnetic field signal is converted into the voltage analog quantity array and transmitted to the processing unit via an analog-to-digital conversion channel.

[0006] Furthermore, in the comprehensive vehicle performance testing method of the present invention, the step of performing tomographic inversion on the voltage analog quantity array to obtain the real-time radial expansion physical deformation of the belt layer inside the tire includes: The voltage analog array is subjected to quadrature demodulation to separate the amplitude characteristic parameter and the phase characteristic parameter; The amplitude characteristic parameter and the phase characteristic parameter are subjected to tomographic inversion in a three-dimensional spatial coordinate system using a regularized inverse problem solving algorithm; Based on the inversion results, a spatial position model of the internal belt layer of the tire during dynamic rotation is constructed, and the physical deformation of the real-time radial expansion is calculated.

[0007] Furthermore, the comprehensive vehicle performance testing method of the present invention, wherein acquiring the acoustic emission signal at the contact interface between the tire and the dynamometer roller to obtain a time-domain signal characterizing the transient viscoelastic state of the contact interface includes: Using a piezoelectric thin film acoustic emission sensor array located beneath the surface coating of the dynamometer roller, high-frequency elastic stress waves released at the moment of a sudden microscopic slippage at the contact interface between the tire and the dynamometer roller are collected. The high-frequency elastic stress wave is converted into a charge fluctuation signal; the charge fluctuation signal is then subjected to signal-to-noise separation and amplitude enhancement to obtain a time-domain signal.

[0008] Furthermore, in the comprehensive vehicle performance testing method of the present invention, the step of performing time-frequency analysis and mechanical mapping on the time-domain signal to obtain the viscoelastic shear tangential stiffness value of the rubber at the contact interface includes: Perform continuous wavelet transform on the time-domain signal to obtain the time-frequency domain multi-scale expansion result; High-frequency energy clusters are extracted from the time-frequency domain multi-scale expansion results, and the dominant frequency domain and attenuation damping coefficient of the energy spectrum are located. By combining the contact mechanics model, the dominant frequency domain and the attenuation damping coefficient are mapped to the viscoelastic shear tangent stiffness value.

[0009] Furthermore, the comprehensive vehicle performance testing method of the present invention, wherein the step of time synchronization and cross-physical field decoupling of the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​to obtain the structural expansion component includes: Interpolation resampling is performed on the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​that are inconsistent with the clock frequency to complete the timestamp synchronization alignment; A dynamic manifold topological space is constructed with the real-time radial expansion physical deformation as the spatial dimension and the viscoelastic shear tangent stiffness value as the time evolution dimension.

[0010] Furthermore, the comprehensive vehicle performance testing method of the present invention, wherein the step of time synchronization and cross-physical field decoupling of the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​to obtain the structural expansion component, further includes: A fractional-order calculus operator is introduced into the coupling state within the topological space of the dynamic manifold to perform orthogonal decomposition in the frequency and spatial domains; viscoelastic softening interference terms are filtered out; and structural expansion components are obtained.

[0011] Furthermore, in the comprehensive vehicle performance testing method of the present invention, the step of generating a speed conversion compensation coefficient based on the structural expansion component includes: A dynamic nonlinear reconstruction function for the true effective rolling radius is constructed based on the structural expansion components. The rotational speed conversion compensation coefficient corresponding to the current sampling period is calculated based on the dynamic nonlinear reconstruction function.

[0012] Furthermore, in the comprehensive vehicle performance testing method of the present invention, the step of compensating the original speed pulse signal fed back by the dynamometer drum using the speed conversion compensation coefficient includes: The speed conversion compensation coefficient is injected into the closed loop of the flux linkage observer of the dynamometer's bottom vector controller; The pulse resampling unit resamples and compensates the phase of the original rotational speed pulse signal according to the updated speed closed-loop reference, so as to output a control pulse that is aligned with the linear velocity of the belt layer inside the tire.

[0013] Secondly, the comprehensive vehicle performance testing system provided by the present invention is applied to the comprehensive vehicle performance testing method as described above, including: The magnetic field signal acquisition module is used to acquire the secondary magnetic field signal formed by the internal belt layer of the tire in response to the excitation magnetic field, and to obtain a voltage analog quantity array characterizing the dynamic spatial displacement of the internal belt layer of the tire. The acoustic signal acquisition module is used to acquire the acoustic emission signal at the contact interface between the tire and the dynamometer roller, and obtain a time-domain signal characterizing the transient viscoelastic state of the contact interface. The tomographic inversion module is used to perform tomographic inversion on the voltage analog quantity array to obtain the real-time radial expansion physical deformation of the belt layer inside the tire. The time-frequency analysis module is used to perform time-frequency analysis and mechanical mapping on the time-domain signal to obtain the viscoelastic shear tangential stiffness value of the rubber at the contact interface. The synchronization decoupling module is used to synchronize the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​in time and decouple them across physical fields to obtain the structural expansion components. The speed compensation module is used to generate a speed conversion compensation coefficient based on the structural expansion component, and to use the speed conversion compensation coefficient to compensate the original speed pulse signal fed back by the dynamometer roller, so as to correct the deviation between the linear velocity of the dynamometer roller and the actual wheel-side speed of the vehicle.

[0014] Beneficial effects of this invention: The comprehensive vehicle performance testing method and system provided by this invention effectively overcomes the limitations of traditional testing technologies that rely solely on static geometric parameters for speed conversion through a cross-physical field fusion mechanism of magnetic field detection and acoustic sensing. The secondary magnetic field signal captured by the magnetic field signal acquisition module can penetrate the rubber medium and directly reflect the spatial displacement law of the tire's internal belt layer. Combined with the real-time radial expansion physical deformation calculated by tomographic inversion technology, it provides precise geometric dimension support for quantifying the structural stretching of the tire under high-speed centrifugal force. The synchronously introduced acoustic signal acquisition module captures high-frequency elastic stress waves at the contact interface and uses time-frequency analysis and mechanical mapping to obtain the viscoelastic shear tangent stiffness value, characterizing the change in contact characteristics caused by temperature rise and softening from the perspective of material mechanics evolution. By constructing a dynamic manifold topological space through a synchronous decoupling module and introducing fractional-order calculus operators, this invention can accurately isolate the rubber viscoelastic time-delay interference and extract the structural expansion component characterizing the true radial outward expansion trend of the tire skeleton. Based on the real-time reconstruction of the true effective rolling radius and the generated speed conversion compensation coefficient using the structural expansion component, the flux linkage observer closed loop of the dynamometer's underlying vector controller is directly applied, achieving dynamic alignment between the dynamometer roller pulse signal and the vehicle's actual wheel-side speed at the physical linear velocity level. This compensation mechanism, based on multi-source physical information fusion, fundamentally eliminates the problem of inconsistent sampling timing caused by tire nonlinear deformation under high dynamic conditions, significantly improving the objectivity and accuracy of the vehicle's dynamic characteristic quantity assessment under extreme operating boundaries. Attached Figure Description

[0015] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on the drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating the comprehensive vehicle performance testing method of the present invention. Detailed Implementation

[0017] To make the technical solution of the present invention clearer, the present invention will be clearly and completely described below with reference to specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. The present invention provided by various embodiments will be described in detail below with reference to the accompanying drawings. To better understand the purpose of the present invention, the present invention will be described in further detail below.

[0018] In a first aspect, the comprehensive performance testing method for automobiles provided by the present invention includes: Step 1: Obtain the secondary magnetic field signal generated by the excitation magnetic field in response to the belt layer inside the tire, and obtain a voltage analog quantity array characterizing the dynamic spatial displacement of the belt layer inside the tire. Step 2: Obtain the acoustic emission signal at the contact interface between the tire and the dynamometer roller to obtain a time-domain signal characterizing the transient viscoelastic state of the contact interface. Step 3: Perform tomographic inversion on the voltage analog array to obtain the real-time radial expansion physical deformation of the belt layer inside the tire; Step 4: Perform time-frequency analysis and mechanical mapping on the time-domain signal to obtain the viscoelastic shear tangent stiffness value of the rubber at the contact interface. Step 5: Time synchronization and cross-physical field decoupling are performed on the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​to obtain the structural expansion components; Step 6: Generate a speed conversion compensation coefficient based on the structural expansion component, and use the speed conversion compensation coefficient to compensate the original speed pulse signal fed back by the dynamometer roller, so as to correct the deviation between the linear velocity of the dynamometer roller and the actual wheel-side speed of the vehicle.

[0019] In this invention, "real-time radial expansion physical deformation" can be understood as the dynamic displacement along the radial direction of the tire relative to the static initial geometric profile when the tire is rotating; "viscoelastic shear tangent stiffness value" can be understood as the local dynamic equivalent stiffness parameter of the tire rubber contact interface under instantaneous shear; "structural expansion component" can be understood as the expansion component mainly caused by the centrifugal stretching of the tire skeleton structure, separated from the overall radial change of the tire after time synchronization and cross-physical field decoupling; "speed conversion compensation coefficient" can be understood as a dynamic proportional factor used to correct the deviation between the original speed pulse of the dynamometer roller and the actual wheel-side speed of the vehicle.

[0020] like Figure 1 As shown, this embodiment provides a method for comprehensive vehicle performance testing. During comprehensive vehicle performance testing using a chassis dynamometer, by simultaneously acquiring dynamic spatial displacement information of the tire's internal belt layer and viscoelastic state information of the contact interface between the tire and the dynamometer roller, the effective rolling radius change of the tire under high-speed conditions is dynamically reconstructed. This compensates and corrects the original rotational speed pulse signal fed back by the dynamometer roller, reducing the misalignment between the roller's linear velocity and the vehicle's actual wheel-side rotational speed. The comprehensive vehicle performance testing method provided by this embodiment includes at least the following steps.

[0021] In step 1, the secondary magnetic field signal generated by the excitation magnetic field in response to the belt layer inside the tire is obtained, and a voltage analog array characterizing the dynamic spatial displacement of the belt layer inside the tire is obtained.

[0022] Specifically, a low-frequency alternating magnetic field emitting unit is installed inside the dynamometer drum along the drum's axial or circumferential direction. This unit may include an excitation coil, a drive circuit, and a frequency controller. The emitted low-frequency alternating magnetic field penetrates the outer insulating rubber layer of the tire, forming inductive coupling in the belt layer or steel wire skeleton region inside the tire, and inducing eddy currents on the surface of the conductive steel wire skeleton. Due to radial expansion, local displacement, and dynamic vibration during high-speed tire rotation, the secondary magnetic field generated by these eddy currents changes in amplitude and phase depending on the spatial position of the belt layer. A penetrating giant magnetoresistive (GMR) receiver matrix, located inside or near the drum, acquires the secondary magnetic field and outputs the corresponding voltage signal. To enable subsequent tomographic inversion, the GMR receiver matrix is ​​preferably arranged in a two-dimensional array, with each sensing unit in the array having known spatial coordinates, thus allowing the sampling results to characterize the dynamic positional changes of the belt layer in multiple spatial directions. The acquired analog voltage signals are pre-amplified, filtered, and converted from analog to digital, forming a voltage analog quantity array arranged in chronological order, and then transmitted to the processing unit. The voltage analog array contains both spatial displacement information of the belt layer relative to the transmitting unit and structural dynamic response information under tire rotation.

[0023] In step 2, the acoustic emission signal at the contact interface between the tire and the dynamometer roller is obtained to obtain a time-domain signal characterizing the transient viscoelastic state of the contact interface.

[0024] Specifically, a piezoelectric thin-film acoustic emission sensor array is deployed beneath the wear-resistant coating on the surface of the dynamometer roller. This array can be partitioned along the width of the tire contact strip and the circumferential direction of the roller to detect high-frequency elastic stress waves generated at the tire-roller contact interface during microscopic slippage, instantaneous stick-slip transitions, and localized stress release. When the tire contacts the roller surface under high speed, high normal load, and significant driving force, the rubber material at the contact interface exhibits complex viscoelastic behaviors such as adhesion, shearing, slippage, and recovery, releasing elastic waves during localized abrupt changes. The piezoelectric thin-film acoustic emission sensor array converts these high-frequency elastic stress waves into charge fluctuation signals, which are then processed by a charge amplifier, bandpass filter, and noise suppression module to obtain a time-domain signal reflecting the transient state changes of the contact interface. In some embodiments, adaptive threshold segmentation and background noise modeling can be used to separate the signal and noise of the time domain signal. Specifically, the background environmental noise of the dynamometer during no-load operation is collected in real time, the average noise energy within the time domain sliding window is calculated, and the average noise energy is multiplied by a preset dynamic floating coefficient (such as 1.5 to 2.0) as the adaptive threshold at the current moment to filter out interference signals below the adaptive threshold, thereby suppressing background interference caused by the electric and electromagnetic environment and mechanical vibration of the drum drive. Envelope enhancement or short-time energy amplification can also be used to improve the identification of weak events.

[0025] In step 3, the voltage analog array is subjected to tomographic inversion to obtain the real-time radial expansion physical deformation of the belt layer inside the tire.

[0026] Specifically, the processing unit first performs orthogonal demodulation on the voltage analog signal array to decompose the original AC response in each sampling channel into in-phase and quadrature components, and further extracts amplitude and phase characteristic parameters. The amplitude characteristic parameter mainly reflects the change in local magnetic coupling strength, while the phase characteristic parameter mainly reflects the signal propagation path and spatial position changes. Subsequently, a regularized inverse problem solving algorithm is used to perform tomographic inversion of the amplitude and phase characteristic parameters in a three-dimensional spatial coordinate system. The regularized inverse problem solving algorithm can employ Tikhonov regularization, truncated singular value decomposition, sparse constraint reconstruction, or other methods suitable for solving underdetermined inverse problems. Its purpose is to recover the spatial position distribution of the belt layer during rotation based on the multi-channel response of the receiver matrix. Based on the inversion results, a spatial position model of the belt layer inside the tire during dynamic rotation is established, and this spatial position model is compared with the static initial geometric model of the tire to calculate the real-time radial expansion physical deformation at the current sampling time.

[0027] In one optional implementation, the tire circumference can be divided into multiple discrete angular units. The change in relative static radius of each angular unit is calculated, and then, through weighted averaging, local peak extraction, or contact zone filtering, a real-time radial expansion physical deformation variable characterizing the current trend of the tire's effective rolling radius is obtained. This real-time radial expansion physical deformation variable can be represented by either absolute displacement values ​​or a relative radius change rate.

[0028] In step 4, time-frequency analysis and mechanical mapping are performed on the time-domain signal to obtain the viscoelastic shear tangential stiffness value of the rubber at the contact interface.

[0029] Specifically, the processing unit performs continuous wavelet transform on the time-domain signal obtained in step 2 to obtain the time-frequency domain multi-scale expansion results. Compared with the Fourier analysis method with a fixed window length, continuous wavelet transform is more suitable for characterizing the high-frequency energy distribution of transient events at the contact interface at different time scales. Subsequently, high-frequency energy clusters are extracted from the time-frequency domain multi-scale expansion results, and the corresponding dominant frequency domain and the attenuation damping coefficient related to the signal envelope attenuation are identified in the energy spectrum. Among them, the dominant frequency domain can reflect the characteristic frequency band of the interface elastic release event, and the attenuation damping coefficient can reflect the energy consumption degree of the rubber material in the contact shear process. Further combined with the contact mechanics model, the dominant frequency domain and the attenuation damping coefficient are mapped to viscoelastic shear tangent stiffness values. The contact mechanics model can be established based on the equivalent Kelvin-Voigt model, the generalized Maxwell model, or the empirical fitting model of tire rubber material, and is used to convert the time-frequency parameters in the acoustic emission response into mechanically meaningful equivalent stiffness parameters. The viscoelastic shear tangent stiffness values ​​obtained thereby can be used to characterize the softening degree, shear bearing capacity, and material response state of the tire contact interface at the current sampling time.

[0030] In step 5, the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​are synchronized in time and decoupled across physical fields to obtain the structural expansion component.

[0031] Specifically, since the sampling frequencies, trigger clocks, and signal delays of the magnetic field sampling link and the acoustic emission sampling link may differ, the real-time radial expansion physical deformation and viscoelastic shear tangent stiffness values ​​are first subjected to timestamp unification and interpolation resampling. In some embodiments, linear interpolation, cubic spline interpolation, or a time window-matching-based resampling method can be used to align the two heterogeneous data sequences on a unified time axis. After time synchronization is completed, a dynamic manifold topology space is constructed with the real-time radial expansion physical deformation as the spatial dimension and the viscoelastic shear tangent stiffness value as the temporal evolution dimension. This dynamic manifold topology space is used to characterize the coupling relationship between the tire skeleton structure response and the rubber contact interface response.

[0032] Furthermore, fractional-order calculus operators are introduced into the coupled states within the dynamic manifold topology to model the coupling relationships exhibiting memory and viscoelastic time-delay characteristics. Orthogonal decomposition in the frequency and spatial domains is then performed to separate the interference terms caused by rubber viscoelastic softening and the structural terms caused by the centrifugal expansion of the belt layer from the comprehensive response. Since the effective rolling radius of the tire changes at high speeds, it is affected by both the centrifugal stretching of the internal steel wire skeleton and the softening and local collapse of the rubber in the contact area. Therefore, it is difficult to directly distinguish the true structural expansion components using only a single physical quantity. This invention, through cross-physics joint modeling, filters out the viscoelastic softening interference terms to obtain structural expansion components more suitable for characterizing the true radial outward expansion trend of the tire skeleton. These structural expansion components can be scalars or vectorized parameters distributed along the tire circumference or contact zone region.

[0033] In step 6, a speed conversion compensation coefficient is generated based on the structural expansion component, and the original speed pulse signal fed back by the dynamometer drum is compensated using the speed conversion compensation coefficient.

[0034] Specifically, the processing unit constructs a dynamic nonlinear reconstruction function for the true effective rolling radius based on the structural expansion component. This dynamic nonlinear reconstruction function describes the relationship between the tire's actual effective rolling radius and its static nominal radius under the current operating conditions. This function can be implemented using a lookup table combined with interpolation, a piecewise nonlinear function, a state-space estimation function, or an empirical fitting function, as long as it can output the corresponding effective rolling radius estimate in real time based on the structural expansion component. Subsequently, the speed conversion compensation coefficient corresponding to the current sampling period is calculated based on the dynamic nonlinear reconstruction function. Preferably, the speed conversion compensation coefficient can be characterized as the ratio between the static nominal rolling radius and the real-time effective rolling radius, or as a dynamic correction coefficient between the original roller pulse frequency and the target wheel-side pulse frequency.

[0035] Furthermore, the speed conversion compensation coefficient is injected into the closed loop of the flux linkage observer in the dynamometer's bottom-level vector controller to update the speed closed-loop reference, making the dynamometer control system's estimation of the current vehicle wheel-side speed more consistent with the actual tire rotation state. Then, the pulse resampling unit performs resampling and phase compensation on the original speed pulse signal based on the updated speed closed-loop reference, outputting a control pulse aligned with the linear velocity of the tire's internal belt layer. The compensated control pulse can be used by the dynamometer's speed conversion, dynamic analysis, load control, and test result output modules, thereby correcting the deviation between the dynamometer roller linear velocity and the vehicle's actual wheel-side speed, improving the accuracy of dynamic characteristic acquisition under extreme operating conditions.

[0036] In this embodiment, the acquisition and tomographic inversion process of the voltage analog quantity array are further explained.

[0037] In some embodiments, the excitation frequency of the low-frequency alternating magnetic field transmitting unit is set to a frequency band capable of penetrating the outer rubber layer of the tire and generating a stable induction on the internal conductive steel wire skeleton. A low- to mid-frequency excitation scheme is preferred to balance penetration depth and spatial resolution. Multiple sensing units in the penetrating giant magnetoresistive receiving matrix are distributed in a matrix along the adjacent area of ​​the drum surface, with each sensing unit corresponding to a different spatial sampling position. After receiving the analog voltage output from each channel, the processing unit establishes a sampling dataset that corresponds one-to-one with the geometric position of the receiving matrix.

[0038] During orthogonal demodulation, the voltage of each channel can be phase-locked using a reference excitation signal, and then the in-phase component I and the quadrature component Q can be extracted. Based on these, the amplitude characteristic parameter A and the phase characteristic parameter φ can be calculated. During tomographic inversion, the A and φ of different spatial channels are combined to form an observation vector. Simultaneously, a forward mapping model between the positional change of the belt layer inside the tire and the response of the external receiving matrix is ​​pre-established. By solving the inverse problem between the observation vector and the forward mapping model, the spatial positional distribution of the belt layer at a certain moment can be recovered. Since the tire exhibits continuity during dynamic rotation, in some embodiments, time continuity constraints or smoothing regularization terms can be introduced into the inverse problem solution to improve inversion stability. The final output real-time radial expansion physical deformation can be defined as the average displacement, maximum displacement, or local displacement of the belt layer's outer contour relative to the initial contour along the tire radius direction.

[0039] In this embodiment, the process of acquiring acoustic emission signals at the contact interface and calculating the viscoelastic shear tangent stiffness is further explained.

[0040] A piezoelectric thin-film acoustic emission sensor array can be embedded beneath the coating on the roller surface to reduce the risk of external damage while maintaining high coupling sensitivity to elastic waves at the contact interface. To improve spatial resolution, multiple sensors can be distributed circumferentially to determine local response differences in different contact areas. Before entering the processing unit, the charge fluctuation signal undergoes sequential processing including charge amplification, bandpass filtering, power frequency interference suppression, mechanical resonance suppression, and amplitude enhancement. After performing continuous wavelet transform on the preprocessed time-domain signal, time-frequency energy distribution maps at different scales are obtained. By clustering and peak localization of high-frequency energy clusters, the dominant frequency domain parameters of the current event can be obtained. By performing exponential fitting or piecewise attenuation fitting on the event's trailing envelope curve, the attenuation damping coefficient can be obtained.

[0041] In some embodiments, a calibration model can be pre-established based on different rubber materials, contact pressures, and temperature conditions, relating the dominant frequency domain, damping coefficient, and equivalent shear stiffness. During testing, the real-time extracted dominant frequency domain and attenuated damping coefficient are input into this calibration model to obtain the viscoelastic shear tangential stiffness value at the current sampling moment. If pre-calibration is not used, the equivalent stiffness can also be solved online based on contact mechanics theory and the material dynamic response equation.

[0042] In this embodiment, the time synchronization, cross-physical field decoupling, and compensation coefficient generation processes are further explained.

[0043] Since the sampling period of the voltage analog signal array is usually different from that of the acoustic emission time-domain signal, the two data streams can be labeled with high-precision timestamps and indexed using a unified master clock. After interpolating and resampling the real-time radial expansion physical deformation and viscoelastic shear tangent stiffness values, a first synchronization vector and a second synchronization vector on the same time series are obtained. The first synchronization vector is used as the spatial structure response variable, and the second synchronization vector is used as the material state evolution variable, both embedded in the dynamic manifold topology space.

[0044] In some embodiments, fractional-order calculus operators are used to characterize the history dependence in the response of rubber materials. Orthogonal decomposition can be achieved using principal component separation, independent component separation, time-frequency domain projection separation, or other mathematical methods that can decouple the frequency domain from the spatial domain. For the multiple components obtained after separation, the viscoelastic softening interference term can be removed based on its frequency band distribution, phase hysteresis characteristics, and correlation with the viscoelastic shear tangent stiffness value, retaining only the structural expansion component that mainly corresponds to the stretching of the belt layer structure.

[0045] After obtaining the structural expansion component, the current actual effective rolling radius can be calculated using a dynamic nonlinear reconstruction function. And calculate the speed conversion compensation coefficient based on the speed conversion relationship used by the dynamometer control system. For example, in an optional implementation, if the static nominal scroll radius is... Then you can press The compensation coefficient is generated from its reciprocal form, and the specific value can be set according to the conversion interface definition of the underlying controller. After the compensation coefficient is written into the closed loop of the flux linkage observer, the pulse resampling unit reconstructs the target pulse frequency according to the updated speed closed loop reference, and matches the output control pulse with the linear velocity of the belt layer inside the tire through phase compensation, thereby reducing the roller speed measurement error caused by the dynamic expansion of the tire.

[0046] This invention also provides a comprehensive vehicle performance testing system, applicable to the comprehensive vehicle performance testing method described in any of the above embodiments. The system includes a magnetic field signal acquisition module, an acoustic signal acquisition module, a tomography inversion module, a time-frequency analysis module, a synchronization decoupling module, and a speed compensation module. Each module can be implemented by independent hardware circuits, or by a processor executing program instructions stored in memory in conjunction with sensors and interface circuits.

[0047] The system comprises several modules: a magnetic field signal acquisition module to drive a low-frequency alternating magnetic field transmitting unit and receive the secondary magnetic field response signal output from a penetrating giant magnetoresistive receiver matrix, forming a voltage analog array characterizing the dynamic spatial displacement of the belt layer; an acoustic signal acquisition module to acquire the charge fluctuation signal output from a piezoelectric thin-film acoustic emission sensor array, perform signal-to-noise separation and amplitude enhancement processing, and output a time-domain signal; a tomography inversion module to perform orthogonal demodulation and three-dimensional tomographic reconstruction based on the voltage analog array, obtaining the real-time radial expansion physical deformation; a time-frequency analysis module to perform continuous wavelet transform on the time-domain signal and calculate the viscoelastic shear tangent stiffness value using a contact mechanics model; a synchronization decoupling module to perform time synchronization, dynamic manifold construction, and cross-physical field decoupling on the above two types of physical quantities, outputting the structural expansion component; and a speed compensation module to generate speed conversion compensation coefficients based on the structural expansion component, and to resample and phase-compensate the original speed pulse signal fed back by the dynamometer drum to correct the deviation between the dynamometer drum linear velocity and the actual wheel-side speed of the vehicle.

[0048] In some embodiments, the processing unit may include an industrial control computer, a digital signal processor (DSP), a field-programmable gate array (FPGA), a microprocessor (MCU), or a combination thereof. The acquisition, quadrature demodulation, and partial tomographic inversion operations of the voltage analog array can be performed by an FPGA or DSP to improve real-time performance; time-frequency analysis, dynamic manifold construction, and dynamic nonlinear reconstruction function calculation can be performed by an industrial control computer or a high-performance processor. The modules can be connected via a high-speed bus, Ethernet bus, or dedicated data interface. The dynamometer's underlying vector controller can receive the speed conversion compensation coefficient via a communication interface and feed back the updated speed closed-loop reference to the pulse resampling unit. This deployment method can meet the real-time processing requirements under highly dynamic testing conditions.

[0049] In some embodiments, the parameters of this invention can be adapted to different vehicle models, tire specifications, roller materials, and test conditions. For example, the excitation magnetic field frequency, giant magnetoresistive receiver matrix density, number of piezoelectric thin-film acoustic emission sensor arrays, resampling period, tomographic inversion regularization weights, continuous wavelet transform scale range, fractional calculus operator order, and dynamic nonlinear reconstruction function form can all be adjusted according to the type of vehicle under test and the test accuracy requirements. For passenger car test scenarios, higher density time-frequency analysis and finer spatial resolution can be used; for commercial vehicle or heavy-duty vehicle test scenarios, the magnetic field penetration capability and contact interface stress wave sampling range can be enhanced. As long as the rotational speed conversion compensation coefficient is still generated based on the cross-physics field fusion mechanism of magnetic field displacement detection and acoustic emission interface state detection, it is considered an embodiment of this invention.

[0050] Secondly, the comprehensive vehicle performance testing system provided by the present invention is applied to the comprehensive vehicle performance testing method as described above, including: The magnetic field signal acquisition module is used to acquire the secondary magnetic field signal formed by the internal belt layer of the tire in response to the excitation magnetic field, and to obtain a voltage analog quantity array characterizing the dynamic spatial displacement of the internal belt layer of the tire. The acoustic signal acquisition module is used to acquire the acoustic emission signal at the contact interface between the tire and the dynamometer roller, and obtain a time-domain signal characterizing the transient viscoelastic state of the contact interface. The tomographic inversion module is used to perform tomographic inversion on the voltage analog quantity array to obtain the real-time radial expansion physical deformation of the belt layer inside the tire. The time-frequency analysis module is used to perform time-frequency analysis and mechanical mapping on the time-domain signal to obtain the viscoelastic shear tangential stiffness value of the rubber at the contact interface. The synchronization decoupling module is used to synchronize the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​in time and decouple them across physical fields to obtain the structural expansion components. The speed compensation module is used to generate a speed conversion compensation coefficient based on the structural expansion component, and to use the speed conversion compensation coefficient to compensate the original speed pulse signal fed back by the dynamometer roller, so as to correct the deviation between the linear velocity of the dynamometer roller and the actual wheel-side speed of the vehicle.

[0051] In the application scenario of high dynamic response chassis dynamometers simulating high-speed cruising and extreme acceleration performance testing of vehicles, the tire, as the core medium for power transmission, is subjected to strong centrifugal force generated by continuous high-speed rotation, combined with the expansion of gas inside the tire due to frictional heat and the softening of material physical properties, causing nonlinear radial displacement of the tire crown. The linear velocity pulses collected by the dynamometer drum cannot directly correspond to the actual rotational state of the driving wheel flange. Conventional speed measurement techniques usually rely on static nominal outer diameter for conversion. Facing extreme testing environments, a multi-dimensional physical state monitoring mechanism must be introduced for underlying intervention. The secondary magnetic field signal formed by the excitation magnetic field in response to the belt layer inside the tire is obtained, resulting in a voltage analog quantity array characterizing the dynamic spatial displacement of the belt layer inside the tire. A low-frequency alternating magnetic field emitting unit located inside the dynamometer drum emits a low-frequency alternating magnetic field that penetrates the outer insulating rubber medium of the tire as an excitation magnetic field to induce eddy currents on the surface of the conductive steel wire skeleton inside the tire. The secondary magnetic field signal derived from the eddy currents is received by a penetrating giant magnetoresistive receiving matrix, and the received secondary magnetic field signal is converted into a voltage analog quantity array and transmitted to the processing unit via an analog-to-digital conversion channel.

[0052] The dynamometer hardware is simultaneously enhanced with an acoustic sensing link to supplement the extraction of contact surface states. Acoustic emission signals from the tire-dynamometer roller contact interface are acquired to obtain a time-domain signal characterizing the transient viscoelastic state of the interface. Using a piezoelectric thin-film acoustic emission sensor array positioned beneath the dynamometer roller's surface coating, high-frequency elastic stress waves released at the moment of abrupt microscopic slippage at the tire-dynamometer roller contact interface are collected. These high-frequency elastic stress waves are converted into charge fluctuation signals, which are then subjected to signal-to-noise separation and amplitude enhancement to obtain the time-domain signal.

[0053] The multi-channel acquired underlying electrophysical signals then enter the decoding and computation stage. A tomographic inversion is performed on the voltage analog signal array to obtain the real-time radial expansion physical deformation of the tire's internal belt layer. Orthogonal demodulation is then performed on the voltage analog signal array to separate the amplitude and phase characteristic parameters. A regularized inverse problem solving algorithm is used to perform a tomographic inversion of the amplitude and phase characteristic parameters in a three-dimensional spatial coordinate system. Based on the inversion results, a spatial position model of the tire's internal belt layer during dynamic rotation is constructed, and the real-time radial expansion physical deformation is calculated.

[0054] The analysis of the acoustic data channel focuses on frequency shift characteristics and energy dissipation patterns. Time-frequency analysis and mechanical mapping are performed on the time-domain signal to obtain the viscoelastic shear tangent stiffness of the rubber at the contact interface. Continuous wavelet transform is performed on the time-domain signal to obtain multi-scale expansion results in the time-frequency domain. High-frequency energy clusters are extracted from the multi-scale expansion results in the time-frequency domain, and the dominant frequency domain and attenuation damping coefficient of the energy spectrum are located. Combined with the contact mechanics model, the dominant frequency domain and attenuation damping coefficient are mapped to viscoelastic shear tangent stiffness values.

[0055] Data streams acquired by heterogeneous sensors in physical space inevitably exhibit asynchronous misalignment in terms of time reference. To address this, time synchronization and cross-physics field decoupling are performed on the real-time radial expansion physical deformation and viscoelastic shear tangent stiffness values ​​to obtain the structural expansion component. Interpolation resampling is applied to the real-time radial expansion physical deformation and viscoelastic shear tangent stiffness values ​​with inconsistent clock frequencies to achieve timestamp synchronization. A dynamic manifold topology space is constructed with the real-time radial expansion physical deformation as the spatial dimension and the viscoelastic shear tangent stiffness value as the temporal evolution dimension. Fractional-order calculus operators are introduced into the coupled states within the dynamic manifold topology space to perform orthogonal decomposition in the frequency and spatial domains, filtering out viscoelastic softening interference terms to obtain the structural expansion component.

[0056] After separating the deformation component purely caused by the stretching of the skeletal structure, the hardware control reconfiguration process begins. A speed conversion compensation coefficient is generated based on the structural expansion component, and this coefficient is used to compensate the original speed pulse signal fed back from the dynamometer roller, correcting the deviation between the dynamometer roller's linear velocity and the vehicle's actual wheel-side speed. A dynamic nonlinear reconstruction function for the true effective rolling radius is constructed based on the structural expansion component. The speed conversion compensation coefficient corresponding to the current sampling period is calculated based on the dynamic nonlinear reconstruction function. This coefficient is then injected into the closed loop of the flux observer in the dynamometer's underlying vector controller. The pulse resampling unit resamples and performs phase compensation on the original speed pulse signal based on the updated speed closed-loop reference, outputting a control pulse aligned with the linear velocity of the tire's internal belt layer.

[0057] The tire's internal belt layer, acting as a skeletal structure to withstand high-speed centrifugal force and torque, comprises a crisscrossing structure of conductive steel wire cords. A low-frequency alternating magnetic field transmitting unit, deployed inside the dynamometer drum, radiates penetrating electromagnetic waves outward using specially oriented coils. The alternating magnetic field penetrates the outer insulating rubber layer and induces weak eddy currents on the surface of the steel wire skeleton. A penetrating giant magnetoresistive receiving matrix contains hundreds or thousands of miniature magnetic induction nodes, responsible for array-based capture of the secondary magnetic field derived from the evolution of eddy currents. The physical magnetic field signal is digitized into a voltage analog array via an analog-to-digital converter. Specifically, the voltage analog array consists of amplitude characteristic parameters of multi-channel alternating voltage and phase characteristic parameters recording spatial propagation delay. When the wheel expands radially outward under extreme acceleration conditions, the amplitude data at specific coordinate nodes within the voltage analog array shows a significant increase, while the phase characteristic parameters simultaneously exhibit nanosecond-level time delay jumps.

[0058] An array of piezoelectric thin-film acoustic emission sensors is mounted directly beneath the tungsten carbide wear-resistant coating on the dynamometer roller. The sensor contains multiple layers of polyvinylidene fluoride (PVDF) piezoelectric film specifically designed to monitor the high-frequency elastic stress waves released during contact and micro-slippage between the rubber tire tread and the roller. These high-frequency elastic stress waves are converted into microampere-level charge fluctuation signals. These weak charge signals pass through a low-noise preamplifier circuit and a bandpass filter to form an analyzable time-domain signal. This time-domain signal contains a large volume of data, specifically including envelope curve data showing the amplitude evolution over time, timestamp data recording the moments of abrupt changes, and high-frequency pulse sequences carrying the frictional characteristics of the contact surface material. During full-load braking or rapid start-up tests on the chassis dynamic testing platform, the tire contact surface rubber undergoes a dramatic adhesive-slip transition. The acoustic fluctuations captured by the piezoelectric film exhibit dense, high-frequency pulse oscillations within a very short timeframe.

[0059] Real-time radial expansion physical deformation reflects the absolute spatial displacement of the tire crown protruding outward when the wheel is rotating at high speed. The tomographic inversion operation relies on a regularized inverse problem solving algorithm to map the spatial position of the multi-channel voltage analog quantity array. The computation unit calls the forward electromagnetic attenuation model in a three-dimensional spatial coordinate system. The computation process substitutes the observed amplitude and phase characteristic parameters into the model for iterative solution. In each iteration, the absolute value of the residual between the current spatial node coordinate update and the result of the previous iteration is calculated. When the absolute value of the residual is less than a preset convergence tolerance threshold (e.g., ...), ... When the iteration count reaches a preset maximum (e.g., 500 times), the iteration is considered converged and the calculation is stopped to output the final true spatial distribution of the skeleton. The spatial position model contains the three-dimensional coordinate data of thousands of discrete nodes on the surface of the belt layer. The dynamically calculated spatial node coordinates are compared with the nominal geometric contour data of the tire in a stationary state using differential calculation. The numerical matrix obtained from the differential calculation constitutes the real-time radial expansion physical deformation. High dynamic testing requires the measurement and control system to complete tens of thousands of matrix multiplication operations within a millisecond cycle.

[0060] The viscoelastic shear tangent stiffness value characterizes the local equivalent resistance to deformation of rubber materials under high temperature and high shear rate conditions. Time-frequency analysis utilizes continuous wavelet transform to map the one-dimensional time-domain signal into a two-dimensional space containing both time and frequency scales. The multi-scale expansion results in the time-frequency domain intuitively present the energy accumulation state of different frequency components at specific time points. The computation module extracts high-frequency energy clusters representing local micro-fractures and slip abrupt changes from the expansion results. The positioning algorithm further extracts the dominant frequency domain parameters that constitute the largest proportion of the energy spectrum, as well as the damping coefficient reflecting the rate of vibration decay. The dominant frequency domain parameters and the damping coefficient are used as inputs into the equivalent Kelvin-Voigt contact mechanics model. The mechanics model outputs the viscoelastic shear tangent stiffness value under specific conditions based on internally preset empirical formulas for rubber mechanics. When the surface temperature of the dynamometer drum rises sharply, the extracted damping coefficient is amplified exponentially.

[0061] The structural expansion component purely refers to the outward expansion displacement caused by the centrifugal stretching of the internal steel wire skeleton. The dynamic manifold topological space is a multi-dimensional mathematical model that integrates spatial position changes with the decay of material mechanical properties. The real-time radial expansion physical deformation variables, after time-stamp alignment, are used as the geometric space dimension. The viscoelastic shear tangent stiffness value is used as the evolution time dimension. Fractional-order calculus operators are used to penetrate the multi-dimensional manifold space and model the viscoelastic coupled response with memory effect. Orthogonal decomposition operations decompose the intertwined composite response in the frequency and spatial domains. The low-frequency, slowly evolving structural stretching characteristics and the high-frequency, violently fluctuating rubber softening characteristics are completely separated under orthogonal decomposition. Viscoelastic softening interference terms representing the surface rubber's thermal softening deformation are filtered out. The extracted core data matrix directly constitutes the structural expansion component.

[0062] The speed conversion compensation coefficient is used to correct the conversion error between the static nominal radius and the actual effective rolling radius after high-speed expansion. The dynamic nonlinear reconstruction function is based on a large amount of test calibration data from a real vehicle dynamometer. The input receives the structural expansion component after removing interference. The reconstruction function internally calls piecewise interpolation logic to deduce the predicted effective rolling radius of the wheel under centrifugal force and torque load. The predicted effective rolling radius is divided by the static wheel radius preset in the chassis dynamometer control system. The calculated floating-point ratio constitutes the speed conversion compensation coefficient. The compensation coefficient is directly injected into the closed-loop loop of the flux linkage observer of the underlying AC power driver. The pulse resampling unit reads the compensation coefficient and changes the duty cycle and phase sequence of the original speed pulse. The hardware-level pulse transmission frequency is recalibrated at the microsecond level.

[0063] After receiving the analog voltage array, the processing unit performs quadrature demodulation to extract signal features. The specific data processing path involves mixing and integrating the alternating voltage signal in the analog voltage array with locally generated in-phase and quadrature reference signals to extract in-phase and quadrature components. Amplitude and phase characteristic parameters are then calculated based on these components. The formulas relied upon for quadrature demodulation are:

[0064]

[0065]

[0066]

[0067] In the formula, Represents in-phase components; Represents the current calculation time; Represents the integral period parameter; Represents the time variable of integration; This represents the alternating voltage signal in the voltage analog quantity array; The angular frequency parameter representing the excitation magnetic field; Represents the in-phase reference signal; Represents orthogonal components; Represents a quadrature reference signal; Representative amplitude characteristic parameter; Represents phase characteristic parameters; This represents the arctangent function. After obtaining the amplitude and phase characteristic parameters, the processing unit uses a regularized inverse problem solving algorithm to perform tomographic inversion in a three-dimensional spatial coordinate system. The specific data processing path is as follows: the amplitude and phase characteristic parameters are combined into an observation vector, and a penalty objective function is constructed using a pre-defined electromagnetic space sensitivity matrix. By minimizing the penalty objective function, the real-time radial expansion physical deformation of the belt layer inside the tire is calculated. The core analytical formula called by the regularized inverse problem solving algorithm is:

[0068] In the formula, This represents the calculated real-time radial expansion physical deformation of the belt layer inside the tire; Represents a pre-defined electromagnetic space sensitivity matrix; The transpose of the predefined electromagnetic spatial sensitivity matrix; This represents the regularization penalty parameter; Represents the regularized smoothing matrix; The transpose of the regularization smoothing matrix; This represents the observation vector composed of a combination of amplitude and phase characteristic parameters; This represents the matrix inversion operator. To address the dynamic value selection process of the regularization penalty parameter, the processing unit introduces a generalized cross-validation algorithm for optimization. The specific processing path is as follows: The processing unit constructs a cross-validation objective function using the observation vectors. The numerator of the cross-validation objective function is the squared L2 norm of the difference between the original observation vector and the forward prediction vector, and the denominator is the squared trace of the difference between the identity matrix and the influence matrix. The influence matrix is ​​the product of the electromagnetic space sensitivity matrix and the regularization smoothing matrix. Within a preset parameter value space, the processing unit iterates through the regularization penalty parameter at a fixed step size to find the parameter value that minimizes the cross-validation objective function. The found parameter value is then used as the optimal regularization penalty parameter for the current sampling period and substituted into the core analytical formula of the tomographic inversion, avoiding model distortion caused by relying on manual experience in parameter selection.

[0069] For the time-domain signal acquired by the piezoelectric thin-film acoustic emission sensor, the processing unit performs a continuous wavelet transform on the time-domain signal. The specific data processing path is as follows: a specific mother wavelet function is selected, and translation and scaling operations are performed on the mother wavelet function. The generated wavelet bases at different scales are then integrated with the time-domain signal through inner product operations. The time-frequency domain multi-scale expansion results are mapped and output, thereby extracting high-frequency energy clusters and locating the dominant frequency domain and attenuation damping coefficient of the energy spectrum. The calculation formula for the continuous wavelet transform is:

[0070] In the formula, Represents the wavelet coefficients in the time-frequency domain multi-scale expansion results; The scaling parameter represents the control of wavelet scaling. The translation parameter represents the control of the wavelet time axis translation; Represents a time variable; This represents the time-domain signal in the acquired time series. The complex conjugate function representing the selected mother wavelet function; Represents the boundary of an infinite integral; This represents the boundary of the negative infinity integral. After extracting the dominant frequency domain and the damping coefficient, the processing unit maps the parameters to viscoelastic shear tangent stiffness values ​​using a contact mechanics model. The specific data processing path is as follows: an equivalent Kelvin-Voigt contact mechanics model is introduced, using the dominant frequency domain as the vibration frequency input and the damping coefficient as the energy loss input. The viscoelastic shear tangent stiffness values ​​of the rubber at the contact interface, reflecting the current temperature and stress state, are calculated using the mechanical transfer function. The mapping calculation formula is:

[0071] In the formula, The value representing the viscoelastic shear tangent stiffness of the rubber at the contact interface; Represents the pre-set and calibrated rubber equivalent contact quality parameters; Represents the constant value of pi; This represents the dominant frequency domain parameters extracted from time-frequency analysis; This represents the attenuation damping coefficient extracted from time-frequency analysis. To eliminate data distortion interference caused by heating and softening, the processing unit introduces fractional-order calculus operators into the coupled states within the dynamic manifold topology and performs orthogonal decomposition. Specifically, for the dynamic deformation process of rubber including memory hysteresis, the fractional-order derivative characteristic sequence of the coupled state variables is calculated, and an orthogonal projection operator matrix is ​​constructed. The mixed deformation variables are projected onto mutually orthogonal elastic structure subspaces and viscoelastic loss subspaces, thereby filtering out viscoelastic softening interference terms and accurately obtaining the structural expansion component entirely generated by centrifugal force. The calculation formulas for the fractional-order derivative characteristic sequence and the orthogonal decomposition projection formulas are as follows:

[0072] In the formula, The order is represented as Fractional calculus operators; This represents the set fractional order parameter; This represents the coupled state variable constructed by fusing real-time radial expansion physical deformation and viscoelastic shear tangent stiffness values. To eliminate differences in physical dimensions, before constructing the coupled state variable, the system pre-maps the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​to Z-score normalization or range normalization methods. The dimensionless interval is then combined according to preset weighting coefficients to form the dimensionless coupled state variables. ; This represents the current calculus calculation time; The independent variable is The gamma function; Represents the integral lag time variable; This represents the first derivative of the integral lag time variable corresponding to the coupled state variable; This represents the structural expansion component that is finally decoupled and separated; The orthogonal projection operator matrix representing the pointer to the flexible structure subspace constructed by the computing system; This represents the real-time radial expansion physical deformation variables within the dynamic manifold topological space before decoupling. Regarding the specific construction logic of the orthogonal projection operator matrix pointing to the elastic structure subspace, the synchronous decoupling module first performs covariance calculations on the real-time radial expansion physical deformation variable sequence and the viscoelastic shear tangent stiffness numerical sequence within a time-domain sliding window to generate a multidimensional state covariance matrix. Subsequently, the synchronous decoupling module performs singular value decomposition on the multidimensional state covariance matrix, separating the left and right singular vector matrices arranged in descending order of eigenvalues. Based on the prior physical assumption that the centrifugal stretching deformation of the tire's internal belt layer has a significantly higher energy proportion than the rubber's softening deformation due to temperature rise, the synchronous decoupling module extracts the first few columns of vectors corresponding to the largest singular values ​​in the left singular vector matrix as a basis vector combination. Then, using the product of the basis vector combination and its transpose, it constructs the orthogonal projection operator matrix pointing to the elastic structure subspace. With this orthogonal projection operator matrix, the system can strictly project the mixed deformation variables onto a spatial dimension coordinate axis that only reflects the actual physical stretching state.

[0073] Based on the decoupled and extracted structural expansion components, the processing unit uses a dynamic nonlinear reconstruction function to reconstruct and deduce the true effective rolling radius, and simultaneously generates speed conversion compensation coefficients. The specific data processing path is as follows: the structural expansion components are substituted into a second-order Taylor expansion fitting polynomial containing first-order linear and second-order nonlinear terms to calculate the true effective rolling radius after stress-tension compensation. Then, the derived true effective rolling radius is divided by the pre-calibrated static nominal rolling radius of the system to calculate the final speed conversion compensation coefficients. The reconstruction deduction and compensation calculation formulas are as follows:

[0074] In the formula, This represents the actual effective rolling radius derived from calculations and deductions; k1 represents the pre-calibrated static nominal rolling radius of the system; k2 represents the first-order linear geometric reconstruction coefficient (dimensionless) determined experimentally; Estr represents the structural expansion component obtained through cross-physics field decoupling operations; k2 represents the second-order nonlinear geometric reconstruction coefficient (dimension is the reciprocal of length) determined experimentally. This represents the final generated speed conversion compensation coefficient used for feedback correction by the underlying equipment.

[0075] The comprehensive vehicle performance testing method provided in this invention involves the construction of multiple core physical and mathematical models and their data processing paths. First, for sensing the spatial position of the belt layer inside the tire, this invention constructs a spatial position model. The spatial position model is based on a pre-defined electromagnetic spatial sensitivity matrix and a static initial geometric model of the tire. In the data processing path, the magnetic field signal acquisition module converts the received secondary magnetic field signal into a voltage analog array and transmits it to the processing unit. The processing unit performs orthogonal demodulation on the voltage analog array, separating the amplitude characteristic parameter reflecting the local magnetic coupling strength and the phase characteristic parameter reflecting the spatial position change. Subsequently, the processing unit uses a regularized inverse problem solving algorithm to map the amplitude and phase characteristic parameters to a three-dimensional spatial coordinate system by solving the inverse problem between the observation vector and the forward mapping model, thereby recovering the spatial node coordinates of the belt layer during rotation. By comparing the real-time recovered spatial node coordinates with the static initial geometric model using differential calculation, the real-time radial expansion physical deformation characteristic of the tire crown's outward protrusion displacement is finally calculated.

[0076] To monitor the material properties of tire contact interfaces, this invention constructs an equivalent Kelvin-Voigt contact mechanics model. Based on the viscoelastic properties of tire rubber materials, this model transforms the interface acoustic emission response parameters into mechanically meaningful equivalent stiffness parameters. In the data processing path, the acoustic signal acquisition module collects high-frequency elastic stress waves through a piezoelectric thin-film acoustic emission sensor array and converts them into time-domain signals. The processing unit performs continuous wavelet transform on the time-domain signals, generating a time-frequency domain multi-scale expansion result containing both time and frequency scales. The computation module extracts high-frequency energy clusters representing local micro-slip from the expansion result and locates the dominant frequency domain parameters and attenuation damping coefficient of the energy spectrum. Finally, the processing unit substitutes the dominant frequency domain as the vibration frequency input and the attenuation damping coefficient as the energy loss input into the equivalent Kelvin-Voigt contact mechanics model, and uses the mechanical transfer function to calculate the viscoelastic shear tangent stiffness value characterizing the current temperature and stress state.

[0077] To achieve deep fusion of multi-source data, this invention constructs a dynamic manifold topological space model. This model uses real-time radial expansion physical deformation as the geometric space dimension and viscoelastic shear tangent stiffness as the evolution time dimension, thus forming a multi-dimensional mathematical manifold characterizing the coupling relationship between the tire skeleton structure response and the rubber contact interface response. In the data processing path, the synchronization decoupling module first performs interpolation resampling on the real-time radial expansion physical deformation and viscoelastic shear tangent stiffness values ​​with inconsistent clock frequencies, completing timestamp synchronization alignment. Subsequently, the processing unit introduces fractional-order calculus operators to model the coupled state with memory effects and viscoelastic time delay characteristics, and performs orthogonal decomposition in the frequency and spatial domains. Through orthogonal projection operations, this invention projects the mixed deformation variables into mutually orthogonal elastic structure subspaces and viscoelastic loss subspaces, thereby filtering out the viscoelastic softening interference term caused by heating and softening, and extracting the structural expansion component purely generated by the centrifugal extension of the skeleton structure.

[0078] Finally, for the closed-loop correction of the speed pulse, this invention constructs a dynamic nonlinear reconstruction function model. This function model is based on a large amount of real-vehicle dynamometer test calibration data and is used to describe the relationship between the actual effective rolling radius of the tire and the static nominal radius. In the data processing path, the processing unit inputs the structural expansion component obtained from cross-physical field decoupling into a second-order Taylor expansion fitting polynomial containing a first-order linear term and a second-order nonlinear term to calculate the true effective rolling radius after stress-tension compensation. The speed compensation module generates a speed conversion compensation coefficient based on the ratio of the true effective rolling radius to the static nominal rolling radius and injects it into the closed loop of the flux observer in the dynamometer's underlying vector controller. The pulse resampling unit resamples and performs phase compensation on the original speed pulse signal based on the updated speed closed-loop reference, thereby outputting a control pulse aligned with the linear velocity of the tire's internal belt layer, correcting the deviation between the dynamometer roller linear velocity and the actual wheel-side speed of the vehicle.

[0079] Embodiment 1 of this invention: In the application scenario of a high dynamic response chassis dynamometer simulating high-speed cruising testing of a conventional sedan, a low-frequency alternating magnetic field emitting unit outputs a magnetic field excitation signal with a center frequency set in the range of 300Hz to 400Hz. The magnetic field excitation signal penetrates the rubber texture of the outer surface of the car tire and induces eddy currents in the internal belt layer region. Simultaneously, a penetrating giant magnetoresistive receiving matrix set inside the dynamometer drum synchronously records the fluctuations of the induced magnetic field as the tire's spatial position changes. The amplitude and phase characteristic parameters generated by the spatial position displacement of the belt layer inside the tire are captured by the analog-to-digital conversion channel and converted into a voltage analog quantity array. The processing unit then calls the Tikhonov regularized inverse problem solving algorithm to perform a tomographic inversion operation on the voltage analog quantity array. The algorithm recovers the three-dimensional spatial coordinate distribution of the belt layer inside the tire during dynamic rotation, allowing the real-time radial expansion physical deformation to be calculated based on the displacement difference between the spatial position model reconstructed from the inversion result and the static model. For monitoring material properties, a piezoelectric thin film acoustic emission sensor array arranged under the surface coating of the dynamometer drum is responsible for capturing high-frequency elastic stress waves at the interface. The time-frequency analysis module performs multi-scale continuous wavelet transform on the acquired time-domain signal. By locating the main frequency domain and attenuation damping coefficient of the high-frequency energy cluster and substituting them into the Kelvin-Voigt contact mechanics model, the viscoelastic shear tangent stiffness value is finally obtained.

[0080] Embodiment 2 of this invention addresses the nonlinear transient deformation problem encountered by high-performance sports cars during extreme acceleration performance testing. Due to the significant radial expansion of the drive wheels under transient high load and high speed, the real-time radial expansion physical deformation exhibits strong dynamic nonlinear evolution characteristics. The synchronous decoupling module first performs interpolation, resampling, and alignment processing on the physical deformation data and stiffness data sequences to meet the timestamp uniformity requirement. The processing unit then establishes a dynamic manifold topology space with the real-time radial expansion physical deformation as the spatial dimension and the viscoelastic shear tangent stiffness value as the temporal evolution dimension. By introducing a fractional-order calculus operator derivative term set to order 0.5, the scheme models the stress hysteresis memory effect of rubber materials and, in conjunction with orthogonal decomposition operations within the manifold topology space, separates the rubber softening interference term caused by temperature rise. The structural expansion component extracted by decoupling is passed as an independent variable to the dynamic nonlinear reconstruction function, thereby guiding the speed conversion compensation coefficient to be calculated in real-time based on the dynamic scaling factor established by the structural expansion component. The compensated conversion ratio is injected into the closed-loop loop of the flux observer of the dynamometer's bottom vector controller, driving the pulse resampling unit to work with the speed closed-loop reference to achieve dynamic correction of the deviation between the dynamometer's drum linear velocity and the actual wheel-side speed.

[0081] Embodiment 3 of this invention: In the application scenario of hill-climbing dynamic testing of large heavy-duty trucks, considering the physical structure characteristics of heavy-duty truck tires with large wall thickness and dense ply layers, the solution increases the driving current power of the low-frequency alternating magnetic field transmitting unit to ensure signal penetration depth. A penetrating giant magnetoresistive receiving matrix adopts a highly integrated arrangement to obtain high-resolution spatial sampling data with a spacing of 5 mm. The magnetic field signal acquisition module simultaneously converts the sensed signal into a voltage analog array characterizing dynamic spatial displacement. The acoustic signal acquisition module uses a bandpass filter component to filter environmental electromagnetic interference noise generated by the dynamometer drive motor, ensuring that the pre-processed time-domain signal can enter the time-frequency analysis module for accurate energy spectrum feature extraction. The viscoelastic shear tangent stiffness value is mapped according to the material's temperature rise softening state reflected by the interface stress wave. Subsequently, the synchronous decoupling module performs cross-physics field fusion calculations on the heterogeneous data sequence within a unified clock cycle and outputs the structural expansion component. The original speed pulse signal fed back by the dynamometer drum is received, the speed conversion compensation coefficient is calculated, and the adjustment is performed in real time with a period of 10 milliseconds. Finally, the control pulse is output by the pulse resampling unit and is strictly aligned with the linear speed of the belt layer inside the tire.

Claims

1. A method for comprehensive vehicle performance testing, characterized in that, include: Step 1: Obtain the secondary magnetic field signal generated by the excitation magnetic field in response to the belt layer inside the tire, and obtain a voltage analog quantity array characterizing the dynamic spatial displacement of the belt layer inside the tire. Step 2: Obtain the acoustic emission signal at the contact interface between the tire and the dynamometer roller to obtain a time-domain signal characterizing the transient viscoelastic state of the contact interface. Step 3: Perform tomographic inversion on the voltage analog array to obtain the real-time radial expansion physical deformation of the belt layer inside the tire; Step 4: Perform time-frequency analysis and mechanical mapping on the time-domain signal to obtain the viscoelastic shear tangent stiffness value of the rubber at the contact interface. Step 5: Time synchronization and cross-physical field decoupling are performed on the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​to obtain the structural expansion components; Step 6: Generate a speed conversion compensation coefficient based on the structural expansion component, and use the speed conversion compensation coefficient to compensate the original speed pulse signal fed back by the dynamometer roller, so as to correct the deviation between the linear velocity of the dynamometer roller and the actual wheel-side speed of the vehicle.

2. The method for comprehensive vehicle performance testing according to claim 1, characterized in that, The process of acquiring the secondary magnetic field signal generated by the excitation magnetic field in response to the internal belt layer of the tire, and obtaining a voltage analog quantity array characterizing the dynamic spatial displacement of the internal belt layer of the tire, includes: A low-frequency alternating magnetic field that penetrates the outer insulating rubber medium of the tire is emitted by a low-frequency alternating magnetic field emitting unit located inside the dynamometer drum as the excitation magnetic field to induce eddy currents on the surface of the conductive steel wire skeleton inside the tire. The secondary magnetic field signal derived from the eddy current is received by a penetrating giant magnetoresistive receiver matrix; the received secondary magnetic field signal is converted into the voltage analog quantity array and transmitted to the processing unit via an analog-to-digital conversion channel.

3. The method for comprehensive vehicle performance testing according to claim 2, characterized in that, The tomographic inversion of the voltage analog array to obtain the real-time radial expansion physical deformation of the belt layer inside the tire includes: The voltage analog array is subjected to quadrature demodulation to separate the amplitude characteristic parameter and the phase characteristic parameter; The amplitude characteristic parameter and the phase characteristic parameter are subjected to tomographic inversion in a three-dimensional spatial coordinate system using a regularized inverse problem solving algorithm; Based on the inversion results, a spatial position model of the internal belt layer of the tire during dynamic rotation is constructed, and the physical deformation of the real-time radial expansion is calculated.

4. The method for comprehensive vehicle performance testing according to claim 3, characterized in that, The step of acquiring the acoustic emission signal at the contact interface between the tire and the dynamometer roller to obtain a time-domain signal characterizing the transient viscoelastic state of the contact interface includes: Using a piezoelectric thin film acoustic emission sensor array located beneath the surface coating of the dynamometer roller, high-frequency elastic stress waves released at the moment of a sudden microscopic slippage at the contact interface between the tire and the dynamometer roller are collected. The high-frequency elastic stress wave is converted into a charge fluctuation signal; the charge fluctuation signal is then subjected to signal-to-noise separation and amplitude enhancement to obtain a time-domain signal.

5. The method for comprehensive vehicle performance testing according to claim 4, characterized in that, The step of performing time-frequency analysis and mechanical mapping on the time-domain signal to obtain the viscoelastic shear tangential stiffness value of the rubber at the contact interface includes: Perform continuous wavelet transform on the time-domain signal to obtain the time-frequency domain multi-scale expansion result; High-frequency energy clusters are extracted from the time-frequency domain multi-scale expansion results, and the dominant frequency domain and attenuation damping coefficient of the energy spectrum are located. By combining the contact mechanics model, the dominant frequency domain and the attenuation damping coefficient are mapped to the viscoelastic shear tangent stiffness value.

6. The method for comprehensive vehicle performance testing according to claim 5, characterized in that, The process of time-synchronizing and decoupling the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​to obtain the structural expansion components includes: Interpolation resampling is performed on the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​that are inconsistent with the clock frequency to complete the timestamp synchronization alignment; A dynamic manifold topological space is constructed with the real-time radial expansion physical deformation as the spatial dimension and the viscoelastic shear tangent stiffness value as the time evolution dimension.

7. The method for comprehensive vehicle performance testing according to claim 6, characterized in that, The step of synchronizing the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​in time and decoupling them across physical fields to obtain the structural expansion components also includes: A fractional-order calculus operator is introduced into the coupling state within the topological space of the dynamic manifold to perform orthogonal decomposition in the frequency and spatial domains; viscoelastic softening interference terms are filtered out; and structural expansion components are obtained.

8. The method for comprehensive vehicle performance testing according to claim 7, characterized in that, The generation of rotational speed conversion compensation coefficients based on the structural expansion components includes: A dynamic nonlinear reconstruction function for the true effective rolling radius is constructed based on the structural expansion components. The rotational speed conversion compensation coefficient corresponding to the current sampling period is calculated based on the dynamic nonlinear reconstruction function.

9. The method for comprehensive vehicle performance testing according to claim 8, characterized in that, The compensation of the original speed pulse signal fed back by the dynamometer drum using the speed conversion compensation coefficient includes: The speed conversion compensation coefficient is injected into the closed loop of the flux linkage observer of the dynamometer's bottom vector controller; The pulse resampling unit resamples and compensates the phase of the original rotational speed pulse signal according to the updated speed closed-loop reference, so as to output a control pulse that is aligned with the linear velocity of the belt layer inside the tire.

10. A comprehensive vehicle performance testing system, applied to the comprehensive vehicle performance testing method as described in any one of claims 1 to 9, characterized in that, include: The magnetic field signal acquisition module is used to acquire the secondary magnetic field signal formed by the internal belt layer of the tire in response to the excitation magnetic field, and to obtain a voltage analog quantity array characterizing the dynamic spatial displacement of the internal belt layer of the tire. The acoustic signal acquisition module is used to acquire the acoustic emission signal at the contact interface between the tire and the dynamometer roller, and obtain a time-domain signal characterizing the transient viscoelastic state of the contact interface. The tomographic inversion module is used to perform tomographic inversion on the voltage analog quantity array to obtain the real-time radial expansion physical deformation of the belt layer inside the tire. The time-frequency analysis module is used to perform time-frequency analysis and mechanical mapping on the time-domain signal to obtain the viscoelastic shear tangential stiffness value of the rubber at the contact interface. The synchronization decoupling module is used to synchronize the real-time radial expansion physical deformation and the viscoelastic shear tangent stiffness values ​​in time and decouple them across physical fields to obtain the structural expansion components. The speed compensation module is used to generate a speed conversion compensation coefficient based on the structural expansion component, and to use the speed conversion compensation coefficient to compensate the original speed pulse signal fed back by the dynamometer roller, so as to correct the deviation between the linear velocity of the dynamometer roller and the actual wheel-side speed of the vehicle.

Images