A super magnetostrictive self-powered sensing integrated tire pressure monitoring system

Abstract

The application discloses a kind of super magnetostrictive self-powered sensing integrated tire pressure monitoring systems, including the vibration capture layer, super magnetostrictive film, plane coil and circuit board sequentially arranged along the same direction;Vibration capture layer is mass-suspension beam composite structure, super magnetostrictive film is attached to its suspension beam base, and circuit board integrates a variety of control and energy storage elements.Vibration capture layer converts tire vibration into super magnetostrictive film mechanical strain, plane coil and super magnetostrictive film form electromagnetic coupling, with energy collection and sensing function, after energy storage reaches standard, switch to sensing mode to detect impedance spectrum, combined with mapping model and BAS-BP neural network to solve tire pressure, compensate interference.The application realizes passive long-term operation, gets rid of battery dependence, significantly reduces measurement error, anti-interference, stable and high precision in full-temperature zone, high temperature resistant and vibration resistant, extremely simple structure, high integration, can be stable and long-term work.

Description

Technical Field

[0001] This invention relates to the field of automotive electronics technology, and in particular to a super magnetostrictive self-powered integrated tire pressure monitoring system. Background Technology

[0002] TPMS (Tire Pressure Monitoring System) is a key component for ensuring vehicle safety and improving fuel economy. Its operational stability and detection accuracy directly affect vehicle driving safety and energy consumption control, making it an important feature in the intelligent and safe development of modern automobiles. Currently, the mainstream tire pressure monitoring technologies in the industry are mainly divided into two categories: ITPMS (Indirect Tire Pressure Monitoring System) and DTPMS (Direct Tire Pressure Monitoring System), which differ significantly in their working principles and detection performance.

[0003] ITPMS does not require a dedicated tire pressure sensor. It relies on the vehicle's existing wheel speed sensors to collect wheel speed signals. Based on the principle that a decrease in tire pressure will lead to a decrease in the tire's rolling radius and a relative increase in wheel speed, it indirectly determines whether the tire pressure is abnormal by comparing the differences in wheel speeds.

[0004] DTPMS, on the other hand, installs pressure sensors inside the tires to directly collect tire pressure information and transmits it wirelessly to the vehicle terminal, enabling real-time tire pressure monitoring and alarms. It is generally superior to ITPMS in terms of detection accuracy, response speed, and reliability.

[0005] Although ITPMS and DTPMS have been widely used, both technologies have inherent defects that restrict the further upgrading and promotion of TPMS.

[0006] ITPMS suffers from significant response lag, triggering alarms only when tire pressure drops significantly or wheel speed differences reach the recognition threshold. It struggles to capture minute changes in tire pressure and fails to provide early warnings of abnormal tire pressure. Furthermore, it cannot effectively identify scenarios where multiple tires are simultaneously underinflated, leading to missed detections. In addition, its detection results are easily affected by factors such as road conditions, load, tire wear, and temperature, resulting in large detection errors and failing to meet the requirements for high-precision monitoring.

[0007] DTPMS is primarily limited by its power supply solution. Sensors are mostly powered by lithium batteries, and battery life determines the overall system lifespan. Replacing batteries increases vehicle usage and maintenance costs. Discarded lithium batteries pose an environmental pollution risk, contradicting environmental development trends. Under harsh conditions such as high temperature, high vibration, and strong centrifugal force, lithium batteries are prone to unstable power supply or even failure, affecting system reliability. Furthermore, existing DTPMS systems have independent energy supply and signal detection modules, resulting in low system integration, complex signal transmission links, and insufficient anti-interference capabilities, further reducing overall system stability.

[0008] In summary, traditional TPMS generally suffers from insufficient detection accuracy, low reliability, high maintenance costs, pollution risks, and reliance on chemical batteries for power. The core root cause lies in the independent energy conversion and information sensing processes, resulting in low energy utilization efficiency. The independent power supply and sensing structures increase system power consumption and circuit complexity, making it difficult to operate stably for extended periods under low power conditions and fundamentally preventing the elimination of dependence on chemical batteries. Therefore, a novel tire pressure monitoring technology is urgently needed to address these issues. Summary of the Invention

[0009] This invention provides a super magnetostrictive self-powered integrated tire pressure monitoring system, which solves the technical problems of traditional TPMS such as reliance on batteries, high maintenance costs, insufficient detection accuracy and reliability, and inability to operate passively. It utilizes the rolling vibration energy of the tire to achieve self-powered operation and realizes intelligent tire pressure monitoring through resonant frequency shift, achieving passive, integrated, highly reliable, and high-precision operation.

[0010] This invention provides a super magnetostrictive, self-powered, integrated tire pressure monitoring system, comprising: The vibration capture layer, the giant magnetostrictive film, the planar coil, and the circuit board are arranged sequentially in the same direction. The vibration capture layer is a mass block-cantilever beam composite structure. The giant magnetostrictive film is attached to the cantilever beam substrate. The circuit board integrates MCU control circuit, analog switch circuit, rectifier energy storage circuit, and supercapacitor. The vibration capture layer converts tire rolling vibration into mechanical strain of the giant magnetostrictive film. The planar coil forms an electromagnetic coupling with the giant magnetostrictive film. The planar coil is a multi-functional multiplexed structure. In power generation mode, it outputs an alternating voltage by inducing changes in magnetic flux. After rectification by the rectifier energy storage circuit, the voltage is stored in a supercapacitor. When the supercapacitor voltage reaches a threshold, the analog switching circuit switches it to sensing mode and inputs a sweep current to detect the coil impedance spectrum. Tire pressure changes alter the tire's radial equivalent stiffness and the system's resonant frequency, causing the resonant peak of the planar coil impedance spectrum to shift. The MCU control circuit extracts the resonant frequency shift and calculates the tire pressure using a tire pressure-frequency mapping model. Interference is compensated through a BAS-BP neural network to achieve passive monitoring.

[0011] Optionally, the super magnetostrictive film is deposited on the cantilever beam substrate of the vibration capture layer using MEMS technology, forming an integrated composite structure with the cantilever beam.

[0012] Optionally, the expression for the tire pressure-frequency mapping model is as follows: , In the above formula, For real-time tire pressure, This is the tire pressure-frequency coupling coefficient. This is the resonant frequency offset. This is a temperature correction factor. This is the temperature compensation amount. This is the system error correction term.

[0013] Optionally, the BAS-BP neural network adopts a 3-10-1 three-layer structure. The input layer contains 3 nodes, corresponding to vibration intensity, ambient temperature, and resonant frequency offset, respectively. The hidden layer contains 10 nodes, which are used to extract nonlinear characteristics, fit data, and separate interference from the input parameters. The output layer contains 1 node, which is used to output the true tire pressure value after nonlinear compensation.

[0014] Optionally, during the training process of the BAS-BP neural network, the objective function is to minimize the prediction error. The input parameters are forward propagated through the hidden layer and the output layer to obtain the predicted tire pressure value. Then, the forward propagation error between the predicted tire pressure value and the actual tire pressure value is calculated. Then, the error is backpropagated layer by layer to each layer of neurons through backpropagation, and the weights and thresholds of each layer are updated synchronously to achieve iterative optimization of network parameters and gradually reduce the detection error.

[0015] Optionally, the system also includes a housing for encapsulating the vibration-capturing layer, the super magnetostrictive film, the planar coil, and the circuit board, providing a mounting base for the system.

[0016] Optionally, the system also includes a first isolation layer and a second isolation layer, which are respectively disposed between the circuit board and the planar coil, and between the planar coil and the super magnetostrictive film, to achieve physical and electrical insulation and avoid electrical crosstalk.

[0017] Optionally, the vibration capture layer, the super magnetostrictive film, the second isolation layer, the planar coil, the first isolation layer, and the circuit board are sequentially and tightly bonded together in the same direction to form a hierarchical integrated structure.

[0018] Optionally, the mass block of the vibration-capturing layer is fixed to the end of the cantilever beam to adjust the equivalent mass of the system and match the tire vibration frequency to improve energy harvesting efficiency.

[0019] Optionally, the cantilever beam of the vibration capture layer is designed as a bending mode, and the system resonant frequency is designed to cover the characteristic frequency band of tire rotation, so as to simultaneously realize the collection of low-frequency vibration energy and the sensing and detection of impedance spectrum.

[0020] One or more technical solutions provided by this invention have at least the following technical effects or advantages: 1. In view of the shortcomings of DTPMS, which rely on lithium battery power supply, have limited lifespan, high maintenance costs, and cause environmental pollution from discarded batteries, this invention achieves passive long-term operation of the system by using tire vibration energy for self-powered operation and supercapacitor energy storage. The lifespan is synchronized with that of the tire, completely eliminating dependence on chemical batteries, reducing the total life cycle cost and meeting environmental protection requirements.

[0021] 2. In response to the shortcomings of ITPMS, such as delayed response, inability to provide early warning, and easy omission of multiple tire underinflation, this invention is based on real-time impedance detection of resonant frequency offset, which can accurately capture minute changes in tire pressure, realize early anomaly warning, support independent monitoring of multiple tires, effectively avoid the risk of missed reporting when multiple tires are underinflated at the same time, and significantly improve the warning reliability of the system.

[0022] 3. To address the shortcomings of ITPMS, which is susceptible to interference from road conditions, load, and temperature, and has large detection errors, this invention adopts resonant frequency offset sensing, co-source energy information channels, and BAS-BP neural network temperature drift compensation technology, which has strong anti-interference capabilities, effectively suppresses temperature drift errors, and achieves high-precision and stable measurement across the entire temperature range.

[0023] 4. In view of the shortcomings of DTPMS in unstable power supply or even failure under high temperature, high vibration and strong centrifugal force conditions, the present invention adopts a battery-free solid structure. The system is resistant to high temperature, vibration and centrifugal force, and operates stably and reliably under extreme conditions.

[0024] 5. To address the shortcomings of traditional TPMS, such as independent energy and sensing modules, redundant structure, low integration, complex circuitry, and high power consumption, this invention adopts a shared magnetostrictive thin film and planar coil magnetic-mechanical-electric coupling interface and time-division multiplexing technology to achieve integrated power generation and sensing. The structure is extremely simple, highly integrated, and has extremely low power consumption, and can operate stably for a long time under low power consumption constraints. Attached Figure Description

[0025] Figure 1 This is a structural diagram of a super magnetostrictive self-powered integrated tire pressure monitoring system according to the present invention; Figure 2 This is a schematic diagram of the integrated mechanical-electrical-information conversion channel of the present invention; Figure 3 This is a schematic diagram of the BAS-BP neural network structure of the present invention; Figure 4This is a distribution diagram of the strong nonlinear coupling characteristic data of tire pressure-frequency-temperature under extreme working conditions according to the present invention; Figure 5 The training convergence curve of the BAS algorithm of this invention for optimizing the parameters of the BP neural network is shown. Figure 6 This is a comparison curve showing the tire pressure calculation accuracy of the traditional BP neural network and the BAS-BP neural network of this invention under wide temperature range and multiple vibration interference conditions. Detailed Implementation

[0026] This invention provides a super magnetostrictive self-powered integrated tire pressure monitoring system, which solves the technical problems of traditional TPMS such as reliance on batteries, high maintenance costs, insufficient detection accuracy and reliability, and inability to operate passively. It utilizes the rolling vibration energy of the tire to achieve self-powering and realizes precise tire pressure sensing through resonant frequency shift, ultimately achieving passive, integrated, highly reliable, and high-precision operation.

[0027] First, the terms appearing in the specification will be explained to facilitate understanding of the technical principles of the present invention.

[0028] Giant magnetostrictive materials are functional materials that can produce significant stretching and deformation under the action of an external magnetic field. Their core working mechanism can achieve changes in internal magnetic flux through the inverse magnetostrictive effect. Specifically, when the tire pressure inside a tire changes, the tire pressure exerts pressure on the giant magnetostrictive material, causing it to undergo minute mechanical deformation. This mechanical deformation changes the magnetic domain structure and arrangement state inside the giant magnetostrictive material, thereby leading to a corresponding change in its permeability. The change in permeability directly causes a change in the internal magnetic flux of the giant magnetostrictive material.

[0029] Bending mode: refers to the inherent and stable bending vibration mode exhibited by a cantilever beam when it resonates under dynamic excitation. Each bending mode corresponds to a definite natural resonant frequency and is the core parameter characterizing the vibration characteristics of elastic components in structural dynamics.

[0030] To better understand, a detailed description will be provided below with reference to the accompanying drawings and specific embodiments. Obviously, the embodiments described in this invention are only a part of the embodiments of this invention, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0031] I. Overall Structure The present invention provides a super magnetostrictive self-powered sensing integrated tire pressure monitoring system (hereinafter referred to as the system). The core of the system is to construct an integrated mechanical-electrical-information coupling conversion logic. The system uses a super magnetostrictive thin film and a planar coil to form a unique physical channel to achieve synchronous conversion of energy and information. Structurally, the system does not distinguish between the power generation unit and the sensing unit. The two share a magneto-mechanical-electric coupling interface, which simplifies the structural redundancy and improves the system integration.

[0032] like Figure 1 As shown, the system includes a housing 1 and various functional components located inside the housing 1. The housing 1 is a rigid protective shell used to encapsulate and protect the internal functional components, while also providing a mounting base for the system and adapting to harsh working environments with high-speed tire rotation and strong centrifugal force.

[0033] The following components are arranged from bottom to top inside the outer casing 1: vibration capture layer 6, super magnetostrictive film 5, second isolation layer 32, planar coil 4, first isolation layer 31, and circuit board 2. The functional components are assembled in sequence and tightly attached to form a hierarchical integrated structure, which ensures the stability of mechanical coupling and electromagnetic coupling and guarantees the reliability of energy conversion and sensing detection.

[0034] Circuit board 2 is fixedly installed inside the bottom of housing 1. Circuit board 2 integrates MCU (Microcontroller Unit) control circuit, analog switch circuit, rectifier energy storage circuit and supercapacitor. The supercapacitor is electrically connected to circuit board 2 and is used to store the electrical energy converted from vibration energy, providing passive power supply support for system operation and completely eliminating the dependence on chemical batteries.

[0035] The first isolation layer 31 is disposed between the circuit board 2 and the planar coil 4 to achieve physical and electrical insulation between the planar coil 4 and the circuit board 2, so as to prevent the circuit signals on the circuit board from interfering with the electromagnetic characteristics of the planar coil and to ensure the stability of energy conversion and signal detection.

[0036] The planar coil 4 is fixedly mounted on the lower surface of the first isolation layer 31, and is arranged vertically and vertically with the supermagnetostrictive thin film 5 to form a compact electromagnetic coupling circuit. The planar coil 4 can be multiplexed in multiple ways by switching the analog switching circuit. In the power generation mode, it senses changes in magnetic flux and outputs electrical energy. In the sensing mode, it inputs a sweep frequency current to detect the impedance spectrum, realizing the multiplexing of energy and information from the same source and through the same channel, further simplifying the system structure.

[0037] The second isolation layer 32 is disposed between the planar coil 4 and the super magnetostrictive film 5 to achieve physical and electrical insulation between the planar coil 4 and the super magnetostrictive film 5, which effectively avoids electrical crosstalk and does not affect the electromagnetic coupling effect between the two, thus ensuring the accuracy of sensing and detection.

[0038] The giant magnetostrictive film 5 is located below the second isolation layer 32. It is coupled to the cantilever beam substrate of the vibration capture layer 6 by surface deposition using MEMS (Micro-Electro-Mechanical Systems) technology, forming an integrated composite structure with the cantilever beam. This ensures that the bending modes of the cantilever beam can be efficiently transmitted to the giant magnetostrictive film 5, causing changes in its magnetic properties and providing a basis for energy conversion and sensing detection.

[0039] Vibration trapping layer 6 is located below the super magnetostrictive film 5 and is a mass block. The cantilever beam composite structure consists of a cantilever beam serving as the supporting base, with a mass block fixed at the end of the cantilever beam. The mass block acts as an inertial element, responsible for efficiently coupling the multidimensional vibrations during tire rotation into the internal conversion channel of the system. The cantilever beam and the giant magnetostrictive film 5 form a reliable mechanical coupling connection. Its core function is to convert the mechanical vibration energy under the tire rolling environment into the strain energy inside the giant magnetostrictive film 5, providing mechanical excitation for subsequent energy conversion and sensing detection.

[0040] The planar coil 4 and the super magnetostrictive thin film 5 form a compact electromagnetic coupling circuit. The planar coil 4 can be multiplexed in multiple ways by switching the analog switching circuit: in the power generation mode, it senses the change in magnetic flux and outputs electrical energy; in the sensing mode, it inputs a sweep frequency current to detect the impedance spectrum, thus realizing the co-source multiplexing of energy harvesting and information sensing.

[0041] The assembly relationship of each functional component is as follows: the outer shell 1 encapsulates and fixes all internal functional components to form an overall protective structure; the circuit board 2 provides circuit control, signal processing and energy storage functions for this system; the planar coil 4 has isolation layers (first isolation layer 31 and second isolation layer 32) on both the upper and lower sides, respectively achieving insulation isolation from the circuit board 2 and the super magnetostrictive film 5; the vibration capture layer 6 (mass block) The cantilever beam is mechanically coupled to the super magnetostrictive film 5 to ensure vibration transmission efficiency.

[0042] Vibration-capturing layer 6 converts tire vibration into strain in the super-magnetostrictive thin film 5. Planar coil 4 forms electromagnetic coupling with the super-magnetostrictive thin film 5. Relying on the inverse magnetostrictive effect of the super-magnetostrictive material, a closed loop of mechanical strain to electrical energy conversion is completed within the same physical channel, together forming a magnetic... machine An electro-coupled structure is used; at the same time, by detecting the dynamic impedance characteristics of the four ends of the planar coil, the resonant frequency offset is extracted, realizing deep integration of sensing and transduction, simplifying the system structure and reducing power consumption.

[0043] The cantilever beam of vibration capture layer 6 is designed for bending modes, with a system resonant frequency target value of several hundred Hz to several thousand Hz, covering the characteristic frequency band of tire rotation, enabling simultaneous collection of low-frequency vibration energy and sensing of impedance spectrum. Specifically, as follows... Figure 2 The diagram shows the integrated mechanical-electrical-information conversion channel of the present invention: the mechanical energy of tire vibration is converted into magnetic energy change through the vibration capture layer 6 and the super magnetostrictive film 5, and then converted into electrical energy through electromagnetic induction of the planar coil 4 and stored in the supercapacitor; when the electrical energy reaches the set threshold, the MCU control circuit is awakened, the impedance spectrum of the planar coil 4 is obtained through frequency sweep excitation, the resonant frequency offset is extracted and the real tire pressure is output after algorithm compensation, so as to realize the same source, same channel and time-division multiplexing of energy harvesting and information perception.

[0044] II. Working Principle 1. Energy Harvesting and Time-Division Multiplexing State Machine This system uses analog switching circuits and timing control to achieve co-source, co-channel, and time-sharing collaborative operation of energy harvesting and information sensing, avoiding mutual interference between the two modes (power generation mode and sensing mode) and improving system reliability and energy utilization.

[0045] Energy harvesting phase: When the vehicle is moving and the speed reaches the activation threshold (e.g., 20 km / h), tire vibration is efficiently coupled and transmitted to the system interior via vibration capture layer 6. Vibration capture layer 6 can significantly improve the transmission efficiency of external vibration to the cantilever beam; vibration excitation acts on the mass block. The cantilever beam composite structure uses a mass block to adjust the system's equivalent mass to reduce the structure's natural frequency, matching the structure's vibration frequency with that of the tire. This induces resonance in the cantilever beam and significantly amplifies the vibration response. The bending modes of the cantilever beam are synchronously transmitted to the giant magnetostrictive film 5, causing strain in the film and triggering magnetic domain reversal. Utilizing the inverse magnetostrictive effect of the giant magnetostrictive material, the magnetic flux inside the film undergoes periodic changes, which in turn induce an alternating voltage in the planar coil 4. This alternating voltage is rectified by the rectifier energy storage circuit on the circuit board 2 and stored in a supercapacitor, completing the energy collection and storage, and providing passive power for subsequent sensing and detection in the system.

[0046] Wake-up and mode switching: When the supercapacitor voltage is charged to the start-up threshold (e.g., 1.8V), the analog switching circuit switches the planar coil 4 from the rectifier energy storage circuit to the frequency sweep excitation circuit, and the system switches from the power generation mode to the sensing mode to start the tire pressure detection process.

[0047] Collaborative Logic: After the sensor detection is completed, the system packages the calculated tire pressure data and sends it to the vehicle ECU (Electronic Control Unit). Then it enters a low-power sleep state, waiting for the supercapacitor to accumulate energy for the next round. This cycle repeats to achieve stable integrated operation without structural redundancy or electromagnetic crosstalk.

[0048] 2. Tire pressure Resonant frequency offset coupling mechanism This system uses "tire pressure" stiffness Vibration transfer function The complete transmission chain of "resonance frequency shift" enables non-contact and accurate monitoring of tire pressure. The core is to utilize the monotonic control effect of tire pressure changes on the system's resonant frequency.

[0049] Physical transmission chain: Changes in tire internal pressure alter the tire's radial equivalent stiffness, which in turn changes the dynamic transmission characteristics and vibration spectrum characteristics of vibration transmitted from the tire to the sensor housing 1; these changes further alter the mass block. cantilever beam The equivalent dynamic parameters of the supermagnetostrictive thin film composite structure ultimately cause a corresponding shift in the system's resonant frequency. The resonant frequency exhibits a clear monotonic correspondence with tire pressure changes: when tire pressure increases, the tire's radial equivalent stiffness increases, and the system's resonant frequency increases accordingly and shifts towards the higher frequency side; when tire pressure decreases, the tire's radial equivalent stiffness decreases, and the system's resonant frequency decreases accordingly and shifts towards the lower frequency side. The tire pressure value can be directly inverted from the resonant frequency shift.

[0050] Frequency sweep identification: The position of the resonant peak of the sensor composite structure shifts synchronously with the external vibration characteristics, and the shift amount has a strictly monotonic mapping relationship with the tire pressure. Under different tire pressures, the peak position of the impedance spectrum of the planar coil 4 shifts. The higher the tire pressure, the more the impedance peak moves towards the higher frequency direction; the lower the tire pressure, the more the impedance peak moves towards the lower frequency direction. The shift amount of the resonant peak can directly characterize the tire pressure change. The MCU control circuit controls the planar coil 4 to perform frequency sweep detection and accurately extracts the resonant frequency shift from the impedance spectrum peak shift.

[0051] Analysis and Calculation: Based on the extracted resonant frequency offset, the MCU control circuit, combined with the built-in tire pressure-frequency mapping model, completes the accurate calculation of tire pressure. This sensing method does not require the pressure-sensitive element to directly contact the gas medium inside the tire. It is a vibration-driven, non-contact tire pressure detection method. Furthermore, the sensing path and energy harvesting path are from the same source, which can significantly improve the system's reliability under harsh conditions such as high-speed rotation and strong centrifugal force, avoiding the shortcomings of contact detection that are susceptible to environmental interference and easy device failure.

[0052] The tire pressure-frequency mapping model is a tire pressure calculation model based on the resonant frequency offset, and its expression is as follows: , In the above formula, For real-time tire pressure, This is the tire pressure-frequency coupling coefficient. This is the resonant frequency offset. This is a temperature correction factor. This is the temperature compensation amount. This is the system error correction term.

[0053] The parameters of this model were obtained through calibration: system resonant frequency data were collected under different tire pressures and ambient temperatures, and determined through data fitting. , , The model parameters are stored in the MCU control circuit to achieve real-time and accurate conversion of resonant frequency offset to tire pressure value.

[0054] 3. Multiphysics Decoupling and Algorithm Correction The magnetic properties of the giant magnetostrictive thin film 5 exhibit significant nonlinearity and are greatly affected by ambient temperature, which can easily lead to errors in tire pressure detection. To address this issue, this system employs the BAS-BP (Beetle Antennae Search-BackPropagation) neural network algorithm for nonlinear compensation, achieving high-precision detection across the entire temperature range.

[0055] like Figure 3 As shown, the BAS-BP neural network adopts a 3-10-1 three-layer structure: the input layer contains 3 nodes, corresponding to vibration intensity, ambient temperature, and resonant frequency offset, respectively; the hidden layer contains 10 nodes, which are used to extract nonlinear characteristics of the input parameters, fit data, and separate interference; the output layer contains 1 node, which is used to output the true tire pressure value after nonlinear compensation.

[0056] During the training of the BAS-BP neural network, the objective function is to minimize the prediction error. The input parameters are forward propagated through the hidden layer and the output layer to obtain the predicted tire pressure value. Then, the forward propagation error between the predicted tire pressure value and the actual tire pressure value is calculated. Then, the error is backpropagated layer by layer to each neuron through backpropagation, and the weights and thresholds of each layer are updated synchronously to achieve iterative optimization of network parameters and gradually reduce the detection error.

[0057] Simultaneously, the BAS (Beetle Antennae Search) algorithm is used to automatically optimize the initial weights and thresholds of the BP (BackPropagation) neural network, providing optimal initial parameters for the neural network, avoiding the BP neural network from getting trapped in local optima, further reducing pressure calculation errors, effectively eliminating interference caused by temperature-induced changes in the permeability of the supermagnetostrictive thin film and vibration intensity, and improving the convergence speed and compensation accuracy of the algorithm.

[0058] Through the above-mentioned multi-physics decoupling and algorithm correction, this system effectively overcomes the technical pain points of large temperature drift and low detection accuracy of traditional magnetostrictive devices, significantly reduces the interference caused by ambient temperature fluctuations, and thus greatly improves the stability and measurement accuracy of the system over a wide operating temperature range, providing a reliable technical guarantee for meeting the high-precision requirements of automotive tire pressure monitoring.

[0059] To further verify the feasibility and practical technical effect of the BAS-BP neural network of the present invention in multi-physics field decoupling, environmental interference suppression and temperature drift compensation, this embodiment conducts system-level simulation experiments and comparative data verification.

[0060] First, a sensor physical mapping model is constructed that takes into account multi-field coupling and extreme operating condition interference. Typical operating conditions for actual automotive use are defined: tire pressure range of 180kPa to 300kPa, and ambient temperature range of -40℃ to 125℃. Random vibration intensity interference and Gaussian noise interference are also superimposed. Because the permeability of the giant magnetostrictive film 5 exhibits a significant second-order nonlinear characteristic with temperature change, and the random noise introduced by vibration is further amplified, a strong nonlinear coupling relationship is formed between the resonant frequency shift acquired by the system and tire pressure and temperature. For example... Figure 4 As shown, under extreme working conditions, the three-dimensional distribution surface of tire pressure-frequency-temperature exhibits obvious distortion and discrete spikes, indicating that conventional linear analytical models are unable to eliminate multi-physics coupling interference and cannot achieve accurate tire pressure calculation. Therefore, nonlinear intelligent compensation algorithms must be introduced to correct errors.

[0061] Based on this, the BAS algorithm is used to globally optimize the initial weights and thresholds of the BP neural network. For example... Figure 5 The figure shows the training convergence curve of the BAS algorithm for optimizing the parameters of the BP neural network according to the present invention. In the initial stage of training, the BAS algorithm rapidly explores the multi-dimensional weight space with its global optimization capability; as the number of iterations increases, the mean square error of the network prediction rapidly decreases and gradually tends to converge smoothly. Experimental results show that the BAS algorithm can effectively overcome the inherent defect of traditional BP neural networks being prone to getting trapped in local optima, providing the optimal initial training parameters for the network model and ensuring the accuracy of subsequent tire pressure data fitting and error compensation.

[0062] At the same time, comparative tests were conducted on the accuracy of tire pressure calculation under wide temperature range and multiple vibration interference conditions. For example... Figure 6 The figure shows a comparison curve of the tire pressure calculation accuracy between the traditional BP neural network and the BAS-BP neural network of this invention. Under the test set conditions of applying the same Gaussian vibration noise and temperature drift interference across the entire temperature range of -40℃ to 125℃, the traditional BP neural network without BAS algorithm optimization (… Figure 6 The red cross-shaped scatter plots in the diagram have weak generalization ability and poor resistance to extreme value interference, resulting in large fluctuations in tire pressure prediction errors. The absolute error values ​​of a large number of test samples exceed the high-precision allowable error range of ±10 kPa. However, the BAS-BP neural network of this invention (…) Figure 6 After nonlinear compensation (using the blue circular scattered points in the image), the system can effectively eliminate the coupling interference caused by temperature drift and vibration noise. The tire pressure prediction error of all test samples is stably constrained within the error bandwidth of ±10 kPa, and the detection accuracy and robustness are significantly improved.

[0063] In summary, the simulation comparison test results fully verify that the present invention achieves multi-physics field decoupling and temperature drift and vibration interference collaborative compensation through BAS-BP neural network, which can adapt to the complex service conditions of tires in the full temperature range, strong vibration and multiple interferences, and can stably achieve high-precision and high-reliability detection of passive tire pressure monitoring, thus achieving the expected technical effect.

[0064] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if such modifications and variations fall within the scope of the claims of this invention and their equivalents, then this invention also includes such modifications and variations.

Claims

1. A super magnetostrictive, self-powered, integrated tire pressure monitoring system, characterized in that, include: The vibration capture layer (6), the super magnetostrictive film (5), the planar coil (4) and the circuit board (2) are arranged sequentially in the same direction; the vibration capture layer (6) is a mass block-cantilever beam composite structure, the super magnetostrictive film (5) is attached to the cantilever beam substrate, and the circuit board (2) integrates MCU control circuit, analog switch circuit, rectifier energy storage circuit and supercapacitor; The vibration capture layer (6) converts the rolling vibration of the tire into the mechanical strain of the super magnetostrictive film (5); the planar coil (4) forms an electromagnetic coupling with the super magnetostrictive film (5). The planar coil (4) is a multi-functional multiplex structure. In the power generation mode, the change in magnetic flux is induced to output an alternating voltage. After rectification by the rectifier energy storage circuit, it is stored in the supercapacitor. When the voltage of the supercapacitor reaches the threshold, the analog switch circuit switches it to the sensing mode and inputs a sweep current to detect the coil impedance spectrum. The change in tire pressure changes the radial equivalent stiffness of the tire and the system resonant frequency, causing the resonant peak of the impedance spectrum of the planar coil (4) to shift. The MCU control circuit extracts the resonant frequency shift and calculates the tire pressure by combining the tire pressure-frequency mapping model. The interference is compensated by the BAS-BP neural network to achieve passive monitoring. The BAS-BP neural network adopts a 3-10-1 three-layer structure. The input layer contains 3 nodes, corresponding to vibration intensity, ambient temperature, and resonant frequency offset, respectively. The hidden layer contains 10 nodes, which are used to extract nonlinear characteristics, fit data, and separate interference from the input parameters. The output layer contains 1 node, which is used to output the true tire pressure value after nonlinear compensation. During the training process of the BAS-BP neural network, the objective function is to minimize the prediction error. The input parameters are forward propagated through the hidden layer and the output layer to obtain the predicted tire pressure value. Then, the forward propagation error between the predicted tire pressure value and the actual tire pressure value is calculated. Then, the error is backpropagated layer by layer to each layer of neurons through backpropagation, and the weights and thresholds of each layer are updated synchronously to achieve iterative optimization of network parameters and gradually reduce the detection error.

2. The system according to claim 1, characterized in that, The super magnetostrictive film (5) is deposited on the cantilever beam substrate of the vibration capture layer (6) using MEMS technology, forming an integrated composite structure with the cantilever beam.

3. The system according to claim 1, characterized in that, The expression for the tire pressure-frequency mapping model is as follows: ， In the above formula, For real-time tire pressure, This is the tire pressure-frequency coupling coefficient. This is the resonant frequency offset. This is a temperature correction factor. This is the temperature compensation amount. This is the system error correction term.

4. The system according to claim 1, characterized in that, It also includes a housing (1) for encapsulating the vibration capture layer (6), the super magnetostrictive film (5), the planar coil (4), and the circuit board (2) to provide a mounting base for the system.

5. The system according to claim 1, characterized in that, It also includes a first isolation layer (31) and a second isolation layer (32), which are respectively disposed between the circuit board (2) and the planar coil (4), and between the planar coil (4) and the super magnetostrictive film (5) to achieve physical and electrical insulation and avoid electrical crosstalk.

6. The system according to claim 5, characterized in that, The vibration capture layer (6), the super magnetostrictive film (5), the second isolation layer (32), the planar coil (4), the first isolation layer (31) and the circuit board (2) are sequentially and tightly bonded together in the same direction to form a hierarchical integrated structure.

7. The system according to claim 1, characterized in that, The mass block of the vibration capture layer (6) is fixed to the end of the cantilever beam to adjust the equivalent mass of the system and match the tire vibration frequency to improve energy harvesting efficiency.

8. The system according to claim 1, characterized in that, The cantilever beam of the vibration capture layer (6) is designed as a bending mode, and the system resonant frequency design target value covers the characteristic frequency band of tire rotation, so as to simultaneously realize the collection of low-frequency vibration energy and the sensing and detection of impedance spectrum.

Images