A medical equipment dynamic perception method based on an internet of things

Abstract

This invention discloses a method for dynamic sensing of medical equipment based on the Internet of Things (IoT), belonging to the field of smart medical equipment management technology, to address the problems of difficulty in quantifying the hidden electromechanical losses of medical equipment and the disconnect between high-risk clinical scheduling and the actual situation. First, the positioning trajectory, multi-axis spatial oscillation sequence, and transient power distribution spectrum of the equipment are synchronously collected via IoT terminals. Then, the oscillation sequence is spatially sliced ​​and cross-fused with the power spectrum in a time-frequency manner to separate the electromechanical coupling fatigue stress spectrum. This spectrum is then used to adaptively nonlinearly map and stretch the attenuation baseline, analytically obtaining a dynamic health margin feature vector. Finally, this feature vector is matched with clinical access conditions through multi-level degradation matching, triggering physical explicit locking and unauthorized interception of edge-side electronic tags via an edge gateway. This constructs a closed-loop system from underlying physical sensing to clinical safety scheduling, providing scientific support for the efficient and safe management of medical equipment.

Description

Technical Field

[0001] This invention relates to the field of intelligent medical equipment management technology, specifically to a method for dynamic sensing of medical equipment based on the Internet of Things. Background Technology

[0002] With the deepening of the digital transformation of modern medical systems and the construction of smart hospitals, various high-precision medical equipment plays a core role in clinical diagnosis, treatment, and life support. These medical devices often have the physical characteristics of high-frequency circulation and cross-ward scheduling, and their use directly relates to the life and health safety of patients. In the midst of busy daily medical operations, ensuring the stable operation and precise scheduling of these high-value, highly mobile medical devices is not only a crucial aspect of hospital asset management but also a key infrastructure for maintaining high-quality medical services and preventing medical accidents.

[0003] Existing dynamic sensing solutions for medical equipment primarily rely on linear depreciation assessments based on equipment maintenance records and simple power-on operating time. This single-dimensional monitoring logic completely separates environmental disturbances during the physical transfer of equipment from actual internal hidden losses. In frequent dispatching across wards and floors, precision medical equipment often encounters mechanical shocks such as road bumps and physical collisions. Existing management systems cannot capture the electromechanical coupling damage resulting from the superposition of these mechanical impacts and the equipment's underlying powered operating state. This leads to significant deviations in the accuracy of core probes or internal sensitive electrical components even when the equipment appears undamaged and has not reached its theoretical service life. Consequently, equipment with hidden faults is incorrectly assigned to high-precision emergency care scenarios, severely interfering with the accuracy of clinical diagnosis and treatment and triggering unknown medical risks. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for dynamic sensing of medical equipment based on the Internet of Things, thus solving the problems mentioned above.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for dynamic sensing of medical equipment based on the Internet of Things (IoT), comprising the following steps: S1, using an IoT MEMS sensing terminal fixed at the interface between the medical equipment chassis load-bearing structure and the core probe, synchronously acquiring the indoor positioning trajectory and corresponding multi-axis spatial oscillation sequence of the medical equipment along the physical flow path across wards, and capturing the transient power distribution map of the medical equipment power supply link under the same timestamp reference; S2, spatially slicing the multi-axis spatial oscillation sequence along the indoor positioning flow trajectory of the medical equipment, extracting the continuous envelope oscillation features within each spatial slice, and performing time-domain and frequency-domain cross-fusion of the continuous envelope oscillation features and the transient power distribution map to separate the physical excitation experienced by the medical equipment under powered operation. S3. At the edge computing node, the rated loss cycle curve of the medical equipment is extracted from the equipment attribute library as the attenuation baseline. The attenuation baseline is subjected to adaptive nonlinear mapping stretching by the characteristic amplitude of each frequency band in the electromechanical coupling fatigue stress spectrum. The dynamic health margin feature vector under the current circulation node of the medical equipment is obtained by parsing. S4. The different clinical diagnosis and treatment scenarios for the medical equipment are issued in real time by the clinical equipment resource scheduling platform. The dynamic health margin feature vector is matched with the operating accuracy access conditions in a multi-level downgrade manner to generate the scenario access control instruction. The scenario access control instruction is transmitted to the electronic ink RFID tag on the medical equipment end through the IoT edge gateway to trigger the physical explicit locking of the available clinical scenario permissions and the terminal call authentication interception.

[0006] Furthermore, the specific process of simultaneously collecting the indoor positioning trajectory and corresponding multi-axis spatial oscillation sequence of medical equipment along the physical flow path across wards, and capturing the transient power distribution map of the medical equipment power supply link under the same timestamp reference is as follows: By deploying indoor positioning beacons at the medical equipment flow nodes, the radio frequency positioning tags on the chassis load-bearing structure of the medical equipment are tracked in real time to generate an indoor positioning trajectory with spatial coordinates; the IoT MEMS sensing terminal fixed at the interface between the chassis load-bearing structure and the core probe of the medical equipment is activated, and the three-dimensional acceleration vector sequence of the medical equipment during the pushing motion is extracted according to the preset sampling rate to construct a multi-axis spatial oscillation sequence; by connecting the Hall current sensor and voltage sampling module of the medical equipment power supply bus, the alternating current waveform and transient voltage drop amplitude of the power supply link of the medical equipment under energized operation are extracted and mapped to generate a transient power distribution map; by using the global unified clock pulse of the IoT edge gateway, the indoor positioning trajectory, multi-axis spatial oscillation sequence and transient power distribution map are timestamped and encapsulated.

[0007] Furthermore, the specific process of spatially slicing the multi-axis spatial oscillation sequence along the indoor positioning and circulation trajectory of the medical equipment and extracting the continuous envelope oscillation features within each spatial slice is as follows: The indoor spatial layout map is retrieved, and the indoor positioning trajectory is mapped onto the map, marking the structural abrupt change coordinate nodes where the medical equipment traverses height-change areas; the structural abrupt change coordinate nodes are extracted, a dynamic spatial sliding window is established, and the multi-axis spatial oscillation sequence is truncated and segmented spatially using the dynamic spatial sliding window to generate independent spatial slices; for the multi-axis spatial oscillation sequence within the independent spatial slices, the instantaneous amplitude contour of the multi-axis spatial oscillation sequence is extracted using Hilbert transform, environmental noise components are filtered out, and continuous envelope oscillation features are retained.

[0008] Furthermore, the specific process of separating the electromechanical coupling fatigue stress spectrum of medical equipment under physical excitation during energized operation by performing time-domain and frequency-domain cross-fusion of continuous envelope oscillation characteristics and transient power distribution spectrum is as follows: In the time domain, the peak time of the physical shock wave in the continuous envelope oscillation characteristics is extracted, and the current fluctuation interval and power distortion interval that occur synchronously with the peak time of the physical shock wave in the transient power distribution spectrum are located; in the frequency domain, wavelet packet decomposition is performed synchronously on the continuous envelope oscillation characteristics and the power distortion interval to obtain the physical oscillation energy frequency band and the electrical interference frequency band of the same frequency resonance; a time-frequency cross-spectral density matrix is ​​constructed, and the cross-correlation feature matrix of the physical oscillation energy frequency band and the electrical interference frequency band is reconstructed to separate the mechanical vibration component and the power grid fluctuation component, thus separating the electromechanical coupling fatigue stress spectrum.

[0009] Furthermore, at the edge computing node, the specific process of extracting the rated loss cycle curve of medical equipment from the device attribute library as the attenuation baseline is as follows: Parse the device physical medium access control address broadcast by the IoT MEMS sensing terminal, and construct a low-level attribute addressing link for medical equipment type at the edge computing node; search the device attribute library along the low-level attribute addressing link to extract the set of calibration life thresholds and reference impedance characteristic profiles of electrical components of medical equipment; perform polynomial time series fitting on the set of calibration life thresholds and reference impedance characteristic profiles of electrical components to generate the rated loss cycle curve, and load the rated loss cycle curve onto the edge computing node as the attenuation baseline.

[0010] Furthermore, the specific process of obtaining the dynamic health margin feature vector under the current circulation node of medical equipment by performing adaptive nonlinear mapping stretching on the attenuation baseline using the characteristic amplitudes of each frequency band in the electromechanical coupling fatigue stress spectrum is as follows: Perform frequency band energy integration on the electromechanical coupling fatigue stress spectrum to extract the first and second characteristic amplitudes in the electromechanical coupling fatigue stress spectrum; construct a two-dimensional environmental stress penalty matrix using the first and second characteristic amplitudes; perform tensor convolution operation on the two-dimensional environmental stress penalty matrix and the slope of the current time node of the attenuation baseline to generate a nonlinear stretching mapping factor; perform dynamic compression and amplitude distortion processing on the time axis of the attenuation baseline using the nonlinear stretching mapping factor, extract the baseline end state value after dynamic compression and amplitude distortion processing, and output the dynamic health margin feature vector.

[0011] Furthermore, the specific process of receiving real-time access conditions for the operational precision of medical equipment in different clinical treatment scenarios from the clinical equipment resource scheduling platform, and performing multi-level degradation matching between the dynamic health margin feature vector and the operational precision access conditions to generate scenario access control instructions is as follows: The received operational precision access conditions for medical equipment in different clinical treatment scenarios are parsed into a standardized scenario constraint space containing signal-to-noise ratio thresholds, impedance tolerance boundaries, and immunity levels; the dynamic health margin feature vector is projected into the standardized scenario constraint space, and the multi-dimensional spatial topological distance between the dynamic health margin feature vector and the boundary features of each level of clinical treatment scenarios in the standardized scenario constraint space is calculated; multi-level degradation matching is performed according to the proximity gradient of the multi-dimensional spatial topological distance, mismatched scenario labels are eliminated, the whitelist of available scenarios is retained, and scenario access control instructions are generated.

[0012] Furthermore, the specific process of transmitting scene access control instructions to the e-ink RFID tag on the medical equipment side via the IoT edge gateway, triggering the physical explicit locking of available clinical scene permissions and terminal call authentication interception, is as follows: The scene access control instructions are encapsulated using a communication protocol on the IoT edge gateway side and sent to the microcontroller within the e-ink RFID tag on the medical equipment side; the microcontroller parses the available scene whitelist in the scene access control instructions, drives the display matrix of the e-ink RFID tag to reconstruct the screen, presents the scene graphic identifier, and executes the physical explicit locking of available clinical scene permissions; the microcontroller simultaneously overwrites the available scene whitelist into the storage sector of the e-ink RFID tag, blocks the unauthorized terminal's unauthorized handshake protocol to the storage sector, and establishes the terminal call authentication interception state.

[0013] The present invention has the following beneficial effects: (1) A method for dynamic sensing of medical equipment based on the Internet of Things (IoT) is proposed. This method uses an IoT MEMS sensing terminal fixed to the interface between the medical equipment's chassis load-bearing structure and the core probe to simultaneously collect indoor positioning trajectories, multi-axis spatial oscillation sequences, and transient power distribution maps during the equipment's movement. Spatial slicing is then performed along the positioning trajectory, and the continuous envelope oscillation characteristics and power maps are cross-fused in the time and frequency domains to successfully separate the electromechanical coupling fatigue stress spectrum. This design effectively unifies the mechanical impact environment experienced by the medical equipment during physical movement with the underlying electrical operating load, overcoming the serious defect of traditional equipment management methods that separate the spatial movement environment from the electrical operating state. It achieves high-precision quantification and feature tracing of hidden internal micro-damage induced by vibration and shock in the equipment.

[0014] (2) A method for dynamic sensing of medical equipment based on the Internet of Things (IoT) involves using the separated electromechanical coupling fatigue stress spectrum at the edge computing node to perform adaptive nonlinear mapping stretching on the rated loss baseline of the equipment, obtaining a dynamic health margin feature vector. This vector is then matched with multi-level degradation matching based on the operational precision access conditions of different clinical treatment scenarios, generating control commands that are transmitted through the IoT edge gateway to the end-side electronic ink RFID tag, triggering physical explicit locking and authentication interception. This design directly transforms the underlying physical fatigue sensing data into clinical scheduling access decisions at the business level, and completes tamper-proof closed-loop interception at the equipment's physical terminal. From the physical execution level, it completely eliminates the risk of hidden defective equipment flowing into high-precision clinical scenarios, greatly improving the security and reliability of medical resource scheduling.

[0015] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0016] Figure 1 This is a flowchart of a method for dynamic sensing of medical equipment based on the Internet of Things (IoT) according to the present invention.

[0017] Figure 2 Flowchart for synchronous acquisition and alignment of multimodal underlying sensing data.

[0018] Figure 3 This is a flowchart for separating spatial slices from electromechanical coupling fatigue stress spectrum.

[0019] Figure 4 Flowchart for constructing attenuation baselines and analyzing dynamic health margins.

[0020] Figure 5 This is a flowchart illustrating the logic of multi-level matching and physical terminal security locking. Detailed Implementation

[0021] This application provides an IoT-based method for dynamic sensing of medical equipment, which solves the problem that existing medical equipment management methods cannot accurately and in real time sense the hidden electromechanical coupling damage suffered by equipment during cross-ward transfers, thereby causing clinical scheduling failures and medical safety risks.

[0022] The overall approach of the solution in this application embodiment is as follows: Using IoT sensing terminals deployed at key physical load-bearing points and electrical interfaces of medical equipment, the mechanical oscillation parameters and transient electrical parameters of the power supply link are simultaneously acquired during the three-dimensional spatial displacement process of the equipment. These mechanical and electrical parameters are cross-fused in the spatiotemporal dimension to extract coupled fatigue characteristics that reflect the actual internal electromechanical damage of the equipment. Distributed edge computing nodes are used to perform nonlinear stretching calculations on the extracted fatigue characteristics and the inherent wear baseline of the equipment to obtain dynamic health assessment indicators characterizing the current usability of the equipment. Finally, combined with the different clinical task precision constraints issued by the hospital equipment resource scheduling system, it is determined whether the equipment currently possesses the access qualifications for a specific clinical scenario. The final downgrade matching result and interception command are then sent to the electronic tag on the surface of the equipment for physical lock control display, thereby constructing a complete technical defense line from underlying multimodal perception and edge mapping computing to physical terminal access control.

[0023] Example 1; please refer to Figure 1This invention provides a technical solution: a method for dynamic sensing of medical equipment based on the Internet of Things (IoT), comprising the following steps: S1, using an IoT MEMS sensing terminal fixed at the interface between the medical equipment chassis load-bearing structure and the core probe, synchronously acquiring the indoor positioning trajectory and corresponding multi-axis spatial oscillation sequence of the medical equipment along the physical flow path across wards, and capturing the transient power distribution map of the medical equipment power supply link under the same timestamp reference; S2, spatially slicing the multi-axis spatial oscillation sequence along the indoor positioning flow trajectory of the medical equipment, extracting the continuous envelope oscillation features within each spatial slice, and performing time-domain and frequency-domain cross-fusion of the continuous envelope oscillation features and the transient power distribution map to separate the electromechanical coupling of the medical equipment under physical excitation during energized operation. S3. At the edge computing node, extract the rated loss cycle curve of the medical equipment from the equipment attribute library as the attenuation baseline. Perform adaptive nonlinear mapping stretching on the attenuation baseline through the characteristic amplitude of each frequency band in the electromechanical coupling fatigue stress spectrum, and analyze and obtain the dynamic health margin feature vector under the current circulation node of the medical equipment; S4. Receive the different clinical diagnosis and treatment scenarios for the medical equipment's operating accuracy access conditions issued in real time by the clinical equipment resource scheduling platform. Perform multi-level degradation matching between the dynamic health margin feature vector and the operating accuracy access conditions to generate scenario access control instructions. Transmit the scenario access control instructions to the electronic ink RFID tag on the medical equipment side through the IoT edge gateway to trigger the physical explicit locking of available clinical scenario permissions and terminal call authentication interception.

[0024] In this implementation plan, step S1 primarily functions to achieve high-frequency synchronous acquisition of multi-source underlying physical and electrical data during the transfer of medical equipment across wards. The IoT MEMS sensing terminal refers to a miniature physical sensor manufactured based on microelectromechanical systems (MEMS) technology, capable of accurately capturing the three-dimensional high-frequency minute acceleration changes generated when the equipment chassis is pushed through hospital corridors and over obstacles. The transient power distribution map refers to the set of abrupt waveform changes in the power supply voltage and alternating current of the equipment's internal core probe or motherboard within a millisecond-level time window when the equipment is in a powered-on state. The purpose of this step is to strictly align the data on the mechanical bumps and impacts experienced by the equipment during physical space movement with the electrical fluctuation data of the internal power supply link through a unified system global clock pulse. This overcomes the limitations of conventional medical equipment management, which can only acquire static positions or single on / off states, providing comprehensive underlying multimodal data support for subsequent precise tracing of hidden micro-damage caused by the equipment's movement.

[0025] Step S2's main function is to transform the disordered, low-level collected data into core characteristic indicators that reflect the true physical hidden damage of medical equipment. Spatial slicing refers to the operation of segmenting continuous oscillation signals according to specific physical geographical defense zones such as speed bumps, elevator gaps, and specific thresholds, using indoor positioning coordinates. Electromechanical coupling fatigue stress spectrum refers to a composite characteristic matrix that specifically maps the microscopic electrical distortion and material loss generated synchronously by sensitive connectors or precision electrical components inside medical equipment when subjected to external mechanical excitation, after stripping away the normal stable environmental background noise. The role of this step is to use time-frequency domain cross-fusion technology to jointly analyze the fragmented mechanical vibration and power grid fluctuations, revealing the inherent coupling relationship between physical impact and electrical micro-loss of the core probe, effectively filtering out irrelevant noise interference, and extracting the most critical characteristic spectral lines for quantifying the hidden losses in the circulation environment of high-precision medical equipment.

[0026] Step S3's main function is to use the extracted electromechanical coupling stress characteristics to infer and update the current true remaining health lifespan boundary of medical equipment in real time. Edge computing nodes refer to distributed computing gateways deployed within the hospital ward's local area network, close to the IoT sensing end. They can achieve low-latency, localized data parsing without relying on a large cloud computing cluster. Adaptive nonlinear mapping stretching refers to breaking through the traditional linear amortization depreciation algorithm based on usage time throughout the equipment's lifespan. It directly superimposes and compresses the accelerated aging effect of components caused by environmental stresses such as vibration and shock as a nonlinear, exponential decay variable onto the theoretical operating lifespan baseline set at the factory. This step integrates the equipment's historical static fixed attributes with the current dynamic environmental stress damage to derive a health margin feature vector that objectively reflects the equipment's current detection accuracy and electrical stability, completely eliminating the risk of misjudging the equipment's condition due to a seemingly intact exterior but micro-shifts in internal core components.

[0027] Step S4's main function is to directly translate the actual health status of the equipment derived from the edge side into safe clinical scheduling decisions, and to complete closed-loop access control at the physical terminal of the medical equipment. Multi-level degradation matching refers to the process where, when the dynamic health score of a device falls below the threshold required for high-precision clinical scenarios such as intensive care or emergency resuscitation, the resource scheduling platform automatically allocates it to alternative clinical tasks with lower probe accuracy requirements, such as routine physical examinations in general wards. Electronic ink RFID tags are intelligent interactive nameplates affixed to the outer shell of medical equipment, possessing power-off imaging retention characteristics and an embedded anti-tampering near-field authentication chip. This step aims to connect the entire chain from underlying data perception to high-level business scheduling, and directly transmit the scheduling access results to the physical tags on the equipment surface for status screen locking and display control, while blocking underlying RF communication. From a physical execution level, it forcibly intercepts potentially dangerous equipment lacking resuscitation accuracy from being illegally scanned and accessed, maximizing the absolute security of clinical medical resource allocation.

[0028] Please see Figure 2 Specifically, the process of synchronously collecting indoor positioning trajectories and corresponding multi-axis spatial oscillation sequences of medical equipment along the physical flow path across wards, and capturing the transient power distribution map of the medical equipment power supply link under the same timestamp reference is as follows: Indoor positioning beacons deployed at the medical equipment flow nodes are used to track radio frequency positioning tags on the chassis load-bearing structure of the medical equipment in real time, generating indoor positioning trajectories with spatial coordinates; the IoT MEMS sensing terminal fixed at the interface between the chassis load-bearing structure and the core probe of the medical equipment is activated, and the three-dimensional acceleration vector sequence of the medical equipment during its movement is extracted according to a preset sampling rate to construct a multi-axis spatial oscillation sequence; the alternating current waveform and transient voltage drop amplitude of the power supply link of the medical equipment under energized operation are extracted by the Hall current sensor and voltage sampling module connected to the medical equipment power supply bus, and mapped to generate a transient power distribution map; the indoor positioning trajectory, multi-axis spatial oscillation sequence, and transient power distribution map are timestamped and encapsulated using the global unified clock pulse of the IoT edge gateway.

[0029] In this implementation plan, a synchronous acquisition mechanism for multimodal sensing data streams is constructed for the underlying data acquisition stage during the movement of medical equipment in complex hospital physical environments. Specifically, indoor positioning beacons are deployed at key nodes in hospital corridors and departments. Through periodic radio frequency handshakes with the radio frequency positioning tags on the equipment chassis, the system continuously acquires and maps the spatial coordinate trajectory set of the equipment's movement. This step provides the basic spatial location anchor points for subsequent environmental stress analysis. Simultaneously with acquiring the location trajectory, the system simultaneously activates the IoT MEMS sensing terminal to collect mechanical impact data. The MEMS sensing terminal, or microelectromechanical system sensor, possesses extremely high sensitivity and miniaturization characteristics, enabling it to accurately capture the minute mechanical vibrations generated when the equipment chassis crosses thresholds or speed bumps. The system constructs a multi-axis spatial oscillation sequence by collecting raw acceleration sampling values ​​in three orthogonal directions. To objectively reflect the overall excitation degree of complex structural equipment, the system performs a comprehensive oscillation amplitude calculation using the following formula: The parameters are explained below: Discrete sampling sequence number of mechanical oscillation data; :No. The combined oscillation amplitude of each sampling sequence node; : These correspond to the structural sensitivity weighting coefficients of the device in the three orthogonal physical dimensions of length, width, and height; : Corresponding to the first The system collects the original acceleration samples along three orthogonal dimensions from each sampling sequence node. This calculation step integrates scattered uniaxial vibration data into a unified measure reflecting the overall physical impact intensity. It should be noted that the structural sensitivity weighting coefficient is determined by extracting the strain rate of the medical equipment chassis in each orthogonal direction using finite element analysis and then normalizing it. The higher the strain rate, the higher the weighting coefficient for the load-bearing direction. Simultaneously, to investigate the implicit impact of physical vibration on the underlying electrical state of the equipment, the system uses a non-contact Hall current sensor and a parallel voltage sampling module to acquire power supply link data. When the precision probe inside the medical equipment is subjected to bumps, slight loosening of its connectors can cause transient impedance changes. The system captures this transient characteristic by constructing a transient power distribution spectrum; the calculation formula is as follows: The parameters are explained below: : Discrete sampling sequence number of electrical link data; :No. Transient compensation power characterization value of each sampling sequence node; Transient voltage drop amplitude in the power supply link; Instantaneous value of alternating current waveform; Real-time phase offset angle between voltage and current; High-frequency glitch amplification factor. This calculation step not only calculates the basic instantaneous active power, but also, by introducing an amplification penalty term for the rate of change of current, greatly highlights the transient current spikes and distortion characteristics caused by physical excitation, thereby accurately mapping the microscopic fluctuations in the power consumption of internal components. The method for determining the high-frequency glitch amplification factor is to pre-collect the maximum rate of change of current of this model of medical equipment under static reference operating conditions, and take the reciprocal of the maximum rate of change of current as the high-frequency glitch amplification factor setting value. Finally, since the mechanical oscillation data and electrical power data come from different sensing chips and have their own independent internal clock offsets, the system must use the IoT edge gateway to send a globally unified clock pulse for spatiotemporal alignment. The edge gateway evaluates the transmission delay of each sensor network and performs microsecond-level data interception and alignment judgment. The calculation formula for determining the same timestamp reference is as follows: The parameters are explained below: : The received timestamp of the multi-axis spatial oscillation sequence at the edge gateway; : The received timestamp of the transient power distribution map at the edge gateway; The fixed hardware transmission delay difference between the mechanical sensing link and the electrical sensing link reaching the edge gateway; Synchronization allowable time window threshold. Only data pairs that meet the above tolerance formula are considered valid concurrent features generated by the same physical excitation and are jointly encapsulated. The above synchronization allowable time window threshold is determined by half the period corresponding to the highest sampling frequency among the various sampling modules contained in the system perception network. This ensures the absolute rigor of time domain alignment and prevents coupling fatigue analysis failure caused by heterogeneous data misalignment.

[0030] Please see Figure 3 Specifically, the process of spatially slicing the multi-axis spatial oscillation sequence along the indoor positioning and circulation trajectory of medical equipment and extracting the continuous envelope oscillation features within each spatial slice is as follows: The indoor spatial layout map is retrieved, and the indoor positioning trajectory is mapped onto the map, marking the structural abrupt change coordinate nodes where the medical equipment traverses height-change areas; the structural abrupt change coordinate nodes are extracted, a dynamic spatial sliding window is established, and the multi-axis spatial oscillation sequence is truncated and segmented spatially using the dynamic spatial sliding window to generate independent spatial slices; for the multi-axis spatial oscillation sequence within each independent spatial slice, the instantaneous amplitude contour of the multi-axis spatial oscillation sequence is extracted using Hilbert transform, environmental noise components are filtered out, and continuous envelope oscillation features are retained.

[0031] In this implementation plan, targeting the flow characteristics of medical equipment in complex indoor geographical environments, the system first performs trajectory-based spatial signal slicing and envelope feature extraction. Specifically, the system retrieves an indoor spatial layout map from the hospital's logistics management system via an interface. This map pre-loads three-dimensional structural information for each floor. The system accurately projects the previously collected indoor positioning trajectory onto the coordinate system of this map, comparing and marking structural abrupt change coordinate nodes in areas of height change such as elevator gaps, ward thresholds, or corridor speed bumps during the medical equipment's movement. Since the vibrations of medical equipment moving on flat ground are invalid background data, extracting these structural abrupt change nodes provides precise physical spatial anchors for capturing effective mechanical impacts. Subsequently, the system establishes a dynamic spatial sliding window centered on the structural abrupt change coordinate nodes, truncating and segmenting the lengthy, continuous multi-axis spatial oscillation sequence in spatial dimensions, thereby generating independent spatial slices containing only data from crossing obstacles or collisions. To extract the low-frequency impact profile that truly transmits destructive force into the equipment from the high-frequency, complex mechanical vibrations, the system performs Hilbert transform and noise filtering calculations on the multi-axis spatial oscillation sequence within the independent spatial slices, as follows: The parameters are explained below: : Local sampling sequence index number within an independent spatial slice; :No. Low-frequency continuous envelope oscillation characteristic values ​​corresponding to each sampling sequence node; : The amplitude of the original high-frequency oscillation signal within an independent spatial slice; : Index number of the discrete convolution summation sequence in the Hilbert transform process; The total number of sampling points contained in an independent spatial slice; : Pre-calibrated high-frequency stable ambient noise energy threshold; This step prevents extremely small positive numbers with a denominator of zero. It extracts the instantaneous amplitude profile of the oscillating signal by constructing the imaginary part of the analytical signal, and simultaneously uses an exponential decay term based on a noise threshold to forcibly smooth and filter low-amplitude stable environmental noise components. The method for determining the pre-calibrated high-frequency stable environmental noise energy threshold is as follows: The root mean square (RMS) of the background high-frequency oscillation energy is collected when the medical equipment is stationary on a flat, insulated surface and the internal cooling fan is operating normally. This RMS value is then set as the high-frequency stable environmental noise energy threshold. This calculation process effectively isolates harmless high-frequency equipment shell resonance and accurately preserves the continuous envelope oscillation characteristics characterizing the intensity of physical impact.

[0032] Specifically, the process of performing time-domain and frequency-domain cross-fusion of continuous envelope oscillation characteristics and transient power distribution maps to separate the electromechanical coupling fatigue stress spectrum of medical equipment under physical excitation during energized operation is as follows: In the time domain, the peak time of the physical shock wave in the continuous envelope oscillation characteristics is extracted, and the current fluctuation range and power distortion range that occur synchronously with the peak time of the physical shock wave in the transient power distribution map are located; In the frequency domain, wavelet packet decomposition is performed synchronously on the continuous envelope oscillation characteristics and the power distortion range to obtain the physical oscillation energy frequency band and the electrical interference frequency band of co-frequency resonance; A time-frequency cross-spectral density matrix is ​​constructed, and the cross-correlation feature matrix of the physical oscillation energy frequency band and the electrical interference frequency band is reconstructed to separate the mechanical vibration component and the power grid fluctuation component, thus separating the electromechanical coupling fatigue stress spectrum.

[0033] In this implementation scheme, during the fusion analysis stage of the mechanical oscillation features extracted from spatial slices and the underlying electrical features, the system aims to identify the coupling and destructive factors hidden in complex signals. In the time domain, the system first scans the continuous envelope oscillation features to locate the peak of the physical shock wave, and uses this as a time reference to search for the current fluctuation range and power distortion range occurring within a very short delay before and after this moment in the transient power distribution spectrum. This time-domain alignment screening can initially screen out mechanical and electrical anomalies that are causally related in time. To further quantify this causal relationship, the system simultaneously performs wavelet packet decomposition on the above two sets of intervals in the frequency domain. Wavelet packet decomposition can finely divide the broadband signal into sub-bands of different scales, thereby obtaining the physical oscillation energy band and the electrical interference band resonating at the same frequency. Subsequently, the system constructs a time-frequency cross-spectral density matrix and performs cross-correlation reconstruction and component stripping calculations as follows: The parameters are explained below: Sub-band index number of the physical oscillation energy frequency band; Sub-band index number of the electrical interference frequency band; : Characteristic stress amplitude of electromechanical coupling fatigue stress spectrum under corresponding frequency band coordinates; : The dynamic time sliding window index number in the time-frequency cross-analysis process; The total number of sliding windows divided within the time fluctuation interval corresponding to the peak moment of the physical shock wave; : Wavelet packet energy density of the physical oscillation energy band within the w-th sliding window; Electrical interference frequency band in the first Wavelet packet energy density within a sliding window; : Isolation attenuation coefficient of pure mechanical vibration component; A baseline for purely mechanical vibration energy that does not overlap with electrical fluctuations in time; : Isolation attenuation coefficient of independent power grid fluctuation components; The baseline for independent power grid fluctuation energy uninduced by physical shocks is crucial. The product summation term in the formula, by calculating the energy density product using a time-sliding window, exponentially amplifies mechanical oscillations and electrical distortions that occur simultaneously at the same instant, while suppressing misalignment signals. The subsequent subtraction precisely separates purely electrical fluctuations caused by the instability of the hospital power grid itself from purely mechanical vibrations that do not cause loosening of electrical components. The isolation attenuation coefficients for the purely mechanical vibration component and the independent power grid fluctuation component are determined by applying standard mechanical excitation under power-off conditions and injecting simulated power grid harmonics under completely static conditions, measuring the proportion of residual crosstalk energy under unilateral independent excitation, and mapping the reciprocal of this proportion to the corresponding isolation attenuation coefficients. The resulting electromechanical coupling fatigue stress spectrum, free of all accidental artifacts, becomes the purest and most reliable quantitative indicator for subsequently assessing whether there are hidden electrical micro-damages in the core probes and interfaces of the medical equipment.

[0034] Please see Figure 4 Specifically, at the edge computing node, the process of extracting the rated loss cycle curve of medical equipment from the device attribute library as the attenuation baseline is as follows: Parse the device physical medium access control address broadcast by the IoT MEMS sensing terminal, and construct a low-level attribute addressing link for medical equipment type at the edge computing node; search the device attribute library along the low-level attribute addressing link to extract the set of calibration life thresholds and reference impedance characteristic profiles of electrical components of medical equipment; perform polynomial time series fitting on the set of calibration life thresholds and reference impedance characteristic profiles of electrical components to generate the rated loss cycle curve, and load the rated loss cycle curve onto the edge computing node as the attenuation baseline.

[0035] In this implementation scheme, during the construction phase of the rated wear baseline for medical equipment, the system first needs to accurately identify the physical identity of the monitored equipment. Edge computing nodes, by parsing the device's physical medium access control address periodically broadcast by the IoT MEMS sensing terminal, can uniquely locate the hardware model and communication port of the medical equipment within the complex ward IoT environment, thereby constructing a low-level attribute addressing link directly to the underlying hardware parameters. Along this addressing link, the system retrieves and extracts the set of calibration life thresholds for electrical components and the reference impedance characteristic profile of the specific model of equipment from the equipment attribute database deployed in the hospital. This step aims to obtain the theoretical aging expectation and health impedance benchmark of key components such as the core probe and motherboard capacitors under ideal factory conditions. To transform these discrete factory calibration data into a curve that can be continuously evaluated over time, the system performs polynomial time series fitting on the extracted calibration life thresholds and reference impedance profile, generating the rated wear period curve calculation formula as follows: The parameters are explained below: : Discrete evaluation sequence number of equipment service time; :No. The baseline value of the rated loss period curve corresponding to each sequence node; The highest order of polynomial time series fitting; :No. The natural aging degradation coefficient corresponding to the step; Reference impedance degradation penalty weight; : No. The theoretical values ​​of the factory-set baseline impedance characteristic profiles corresponding to each sequence node. This calculation step constructs a baseline model that reflects the natural aging law of the equipment under no external violent impact by combining high-order polynomials and logarithmic terms. The method for determining the baseline impedance degradation penalty weight is to extract the impedance drift rate of the same batch of medical equipment during accelerated aging destructive testing in a standard laboratory under constant temperature and humidity conditions, and set the normalized mean of the impedance drift rate as the baseline impedance degradation penalty weight. Finally, this rated loss cycle curve is loaded into the memory sandbox of the edge computing node and configured as the attenuation baseline, providing a static reference standard for subsequent dynamic mapping stretching.

[0036] Specifically, the process of obtaining the dynamic health margin feature vector of the medical equipment at the current circulation node by performing adaptive nonlinear mapping stretching on the attenuation baseline using the characteristic amplitudes of each frequency band in the electromechanical coupling fatigue stress spectrum is as follows: Perform frequency band energy integration on the electromechanical coupling fatigue stress spectrum to extract the first and second characteristic amplitudes; construct a two-dimensional environmental stress penalty matrix using the first and second characteristic amplitudes; perform tensor convolution operation between the two-dimensional environmental stress penalty matrix and the slope of the current time node of the attenuation baseline to generate a nonlinear stretching mapping factor; perform dynamic compression and amplitude distortion processing on the time axis of the attenuation baseline using the nonlinear stretching mapping factor, extract the baseline end state value after dynamic compression and amplitude distortion processing, and output the dynamic health margin feature vector.

[0037] In this implementation scheme, during the simulation phase of the equipment's actual health status, the system uses the previously separated electromechanical coupling fatigue stress spectrum to perform nonlinear dynamic correction on the static attenuation baseline. First, the system performs bandgap energy integration on the electromechanical coupling fatigue stress spectrum, extracting a first characteristic amplitude representing the high-frequency mechanical impact damaging the probe, and a second characteristic amplitude representing the low-frequency electrical fluctuations affecting the power supply link. Subsequently, the system constructs a two-dimensional environmental stress penalty matrix using the first and second characteristic amplitudes, and performs a tensor convolution operation between the two-dimensional environmental stress penalty matrix and the tangent slope of the attenuation baseline at the current time node, generating the following formula for calculating the nonlinear stretching mapping factor: The parameters are explained below: : The index of the current dynamic evaluation time; : The nonlinear stretching mapping factor generated at the current moment; : Row direction dimension index of the two-dimensional penalty matrix; : Column-direction dimension index of the two-dimensional penalty matrix; : The maximum number of rows in a two-dimensional penalty matrix; The maximum number of columns in a two-dimensional penalty matrix; : Tensor elements of the two-dimensional environmental stress penalty matrix constructed by discretizing the first characteristic amplitude and the second characteristic amplitude; : Attenuation baseline at The slope gradient of the time node corresponding to the given moment; Electromechanical damage cross-coupling amplification factor; : First characteristic amplitude; Second characteristic amplitude; : Exponential fatigue evolution rate parameter. The purpose of this calculation step is not merely to simply superimpose physical impact and electrical losses, but to transform the chain reaction of mechanical vibration leading to loose electrical interfaces and subsequent short-circuit heating into a penalty factor that expands rapidly with physical stress through tensor convolution and cross-product coupling terms. The method for determining the electromechanical damage cross-coupling amplification factor is as follows: statistical analysis is performed on sample data from historical equipment maintenance records showing internal circuit burnout due to physical drops or bumps. The ratio of the peak physical excitation acceleration to the peak transient current at the time of burnout is extracted, and the statistical median of this ratio is taken as the electromechanical damage cross-coupling amplification factor. After obtaining the nonlinear stretching mapping factor, the system uses the nonlinear stretching mapping factor to perform dynamic compression and amplitude distortion processing on the time axis of the attenuation baseline. The calculation formula is as follows: The parameters are explained below: : The final value of the extracted dynamic health margin feature vector; : The time window variable during the integration process; The theoretical maximum operating service life of medical equipment; The irreversible aging acceleration power exponent is determined by using an integral model to summarize all historically accumulated mapping factors and continuously amplify the penalty effect as the equipment's service time increases, forcibly compressing the remaining length of the baseline. The final state value of the baseline after dynamic compression and amplitude distortion processing is then extracted as the output. The method for determining the aforementioned irreversible aging acceleration power exponent involves using a material fatigue limit tester to obtain the SN fatigue life curve of the core load-bearing component of the equipment, and extracting the absolute value of the fitting slope of this curve in the plastic strain stage as the power exponent. Through this series of complex nonlinear stretching processes, the final dynamic health margin feature vector output by the system perfectly integrates the equipment's static factory quality and the latent damage encountered during dynamic operation, becoming a core benchmark that accurately represents the current usable state of the medical equipment.

[0038] Please see Figure 5Specifically, the process of receiving real-time operational precision access conditions for medical equipment from the clinical equipment resource scheduling platform for different clinical treatment scenarios, and performing multi-level degradation matching between the dynamic health margin feature vector and the operational precision access conditions to generate scenario access control instructions is as follows: The received operational precision access conditions for medical equipment from different clinical treatment scenarios are parsed into a standardized scenario constraint space containing signal-to-noise ratio thresholds, impedance tolerance boundaries, and immunity levels; the dynamic health margin feature vector is projected into the standardized scenario constraint space, and the multi-dimensional spatial topological distance between the dynamic health margin feature vector and the boundary features of each level of clinical treatment scenarios within the standardized scenario constraint space is calculated; multi-level degradation matching is performed according to the proximity gradient of the multi-dimensional spatial topological distance, mismatched scenario labels are removed, the available scenario whitelist is retained, and scenario access control instructions are generated.

[0039] In this implementation plan, for the scheduling and matching stage of medical equipment health status and actual clinical operations, the system first needs to eliminate the semantic gap between the underlying hardware assessment data and the high-level medical business requirements. Specifically, after receiving the business requirements issued by the clinical equipment resource scheduling platform, the system does not use abstract text labels for fuzzy comparison, but instead parses them into a standardized scenario constraint space containing multiple physical dimensions such as signal-to-noise ratio thresholds, impedance tolerance boundaries, and power grid fluctuation immunity levels. This operation precisely quantifies the broad clinical emergency or routine examination requirements into machine-readable underlying hardware performance boundaries. Subsequently, the system projects the dynamic health margin feature vector obtained in the previous steps into this standardized scenario constraint space to accurately assess the equipment's current risk resistance capability and its fit with the requirements of different business scenarios. The system performs a refined admission matching calculation by calculating the multidimensional spatial topological distance between the dynamic health margin feature vector and the boundary features of each level of clinical diagnosis and treatment scenarios within the standardized scenario constraint space. The calculation formula is as follows: The parameters are explained below: Global index number for clinical diagnosis and treatment scenarios; : A multidimensional spatial topological distance evaluation index corresponding to clinical diagnosis and treatment scenarios; The total number of constraint dimensions in the standardized scenario constraint space partitioning; Local index number of the independent constraint dimension; : Importance weight coefficients for corresponding independent constraint dimensions; : Non-linear maximum activation function; The minimum performance admission threshold for the target scenario under the corresponding constraint dimension; The hardware mapping evaluation value of the current dynamic health margin feature vector of medical equipment in the corresponding dimension; : Scaling factor for scenario-based business risk penalties; The inherent high-risk medical level assessment value of the target scenario. The core function of this calculation formula is to achieve one-way penalty interception using a nonlinear maximum activation function. That is, spatial distance accumulation calculation is only performed on the part where the equipment hardware capability is lower than the scenario safety boundary. If the equipment performance is far better than the scenario boundary requirements, the distance increment is zeroed and no reward is given. At the same time, the inherent risk level of the scenario is introduced into the latter half of the formula as a penalty offset term to ensure that the access spatial distance threshold for high-risk scenarios such as intensive care units is extremely strict. The method for determining the above importance weight coefficient is as follows: relying on the hospital's previous medical adverse event operation and maintenance database, the analytic hierarchy process is used to statistically analyze the correlation trigger probability of serious clinical accidents caused by failure of various underlying hardware dimensions of the equipment. After normalizing the extracted correlation trigger probability across the entire range, it is set as the importance weight coefficient of the corresponding constraint dimension. Finally, the system performs multi-level degradation matching based on the proximity gradient of multi-dimensional spatial topological distance. It automatically screens and removes high-risk scenario labels that have excessively large topological distances, meaning that the existing health performance of the equipment can no longer safely support them. The remaining labels that meet the distance safety threshold are packaged and retained as a whitelist of legally usable scenarios, and scenario access control instructions are generated accordingly. This step completely establishes an intelligent decision-making closed loop from the perception of underlying physical damage to equipment to the safe scheduling of high-level medical resources.

[0040] Specifically, the process of transmitting scene access control instructions to the e-ink RFID tag on the medical equipment side via the IoT edge gateway, triggering the physical explicit locking of available clinical scene permissions and the terminal call authentication interception, is as follows: The scene access control instructions are encapsulated using a communication protocol on the IoT edge gateway side and sent to the microcontroller inside the e-ink RFID tag on the medical equipment side; the microcontroller parses the available scene whitelist in the scene access control instructions, drives the display matrix of the e-ink RFID tag to reconstruct the screen, presents the scene graphic identifier, and executes the physical explicit locking of available clinical scene permissions; the microcontroller simultaneously overwrites the available scene whitelist into the storage sector of the e-ink RFID tag, blocks the unauthorized terminal's unauthorized handshake protocol to the storage sector, and establishes the terminal call authentication interception state.

[0041] In this implementation scheme, for the terminal physical execution and hardware-level security locking stages of scheduling decision commands, the system achieves forced control extension from edge cloud logic to end-side physical hardware. First, at the IoT edge gateway, the generated scene access control commands are deeply encapsulated using a low-power wide-area network (LPWAN) communication protocol and then transmitted transparently to the microcontroller within the electronic ink RFID tag attached to the medical equipment. This step leverages the strong anti-interference and building penetration capabilities of IoT low-frequency channels to ensure that even if the medical equipment is moved into an underground warehouse or radiation shielding room with signal blind spots, the authentication control commands can still be delivered reliably with zero packet loss. After receiving and parsing the available scene whitelist from the scene access control commands, the microcontroller immediately drives the display matrix on the surface of the electronic ink RFID tag to reconstruct the pixel image. Due to the extremely low power consumption and power-off image retention characteristics of the electronic ink display substrate, it can clearly and permanently present scene graphic identifiers by controlling the phase transition movement of the electrophoretic particles in the internal microcapsules. This operation directly and explicitly locks the available clinical scenario permissions physically, allowing on-site medical staff to clearly determine whether the equipment is currently downgraded to be limited to routine check-ups or qualified for critical care simply by visually viewing the device tag, without needing any query terminal. This visual interaction eliminates the risk of blind access. To prevent unauthorized operations caused by overlooking the physical identifier, the microcontroller simultaneously converts the available scenario whitelist into hexadecimal encrypted text and overwrites it into the internal near-field communication storage sector of the e-ink RFID tag. When an external medical staff handheld terminal in the clinical department attempts to scan the code or use the medical equipment near-field contact, the authentication microcontroller within the tag executes the underlying verification and authentication interception status calculation formula as follows: The parameters are explained below: : A radio frequency call request command initiated by an external medical handheld terminal for a specific scenario; The discrete output value of the terminal's authentication, interception, and blocking state is: one indicates that the protocol layer interception is initiated, and zero indicates that the data handshake is allowed. Step transition control function; Collision avoidance safety verification margin threshold; : The total number of authorized scenario entries in the available scenario whitelist that are overwritten into the storage sector; : The traversal index sequence number of the whitelist entries for authorized scenarios; The hash ciphertext value of the request scenario features extracted after the request instruction is parsed by the protocol stack; The first pre-burned data in the storage sector The feature hash ciphertext value of a legitimate authorized scenario; The tolerance polarization variance of nonlinear matching in the hash feature space. The method for determining the above-mentioned anti-collision security verification margin threshold is to extract the maximum white noise bit inversion drift rate allowed in the physical layer specification of the underlying near-field communication protocol stack, and combine it with the discrete distribution law of Hamming distance of the hash algorithm to extract the minimum positive probability boundary value under the condition of ensuring that the system has an absolute zero false recognition rate as the security verification margin threshold. This underlying calculation formula accurately determines whether the request scenario exists in the whitelist sequence within microseconds by summarizing the nonlinear Gaussian similarity between the terminal request scenario ciphertext and all legal scenario ciphertexts built into the whitelist. Once it is determined to be an unauthorized forced call, the step transition control function immediately flips the output logic high-level blocking signal, physically cutting off the RF enable pin of the near-field communication chip, thereby blocking the unauthorized handshake protocol of the unauthorized terminal to the storage sector. This not only realizes the physical screen explicitness of the device degradation state, but also builds a strong interception barrier that cannot be bypassed by software at the lowest level of the wireless RF link, ensuring the absolute authority of the medical equipment security scheduling mechanism.

[0042] In summary, this application has at least the following effects: A method for dynamic sensing of medical equipment based on the Internet of Things (IoT) is proposed. By synchronously collecting multi-axis spatial oscillation sequences and transient power distribution maps of medical equipment during cross-ward transfer and performing time-frequency cross-fusion, the electromechanical coupling fatigue stress spectrum is accurately separated. This breaks through the limitation of traditional equipment management that only relies on power-on time to assess depreciation, and realizes the quantitative perception of hidden internal electrical micro-damage induced by physical bumps in high-precision equipment. Furthermore, the stress spectrum is adaptively and nonlinearly mapped and stretched with the inherent attenuation baseline of the equipment using edge computing nodes to obtain a dynamic health margin feature vector that truly reflects the current usability of the equipment. This vector is then matched with the operational precision access conditions of clinical diagnosis and treatment scenarios through multi-level degradation matching. Finally, the electronic ink RFID tag on the medical equipment side is driven by the IoT edge gateway to complete the physical explicit locking of usability scenario permissions and the interception of unauthorized calls. This completely opens up the decision-making closed loop from the underlying multimodal physical damage perception to the high-level clinical resource security scheduling. From the physical execution level, it effectively prevents the medical risk of hidden defective equipment being illegally allocated to high-precision rescue scenarios, greatly improving the precision of smart hospital equipment management and the absolute safety of clinical medical devices.

[0043] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for dynamic sensing of medical equipment based on the Internet of Things, characterized in that, Includes the following steps: S1. Through the IoT MEMS sensing terminal fixed at the interface between the medical equipment chassis load-bearing structure and the core probe, the indoor positioning trajectory and corresponding multi-axis spatial oscillation sequence of the medical equipment under the physical flow path across wards are collected simultaneously, and the transient power distribution spectrum of the medical equipment power supply link under the same timestamp reference is captured. S2. The multi-axis spatial oscillation sequence is spatially sliced ​​along the indoor positioning and circulation trajectory of the medical equipment. The continuous envelope oscillation features in each spatial slice are extracted. The continuous envelope oscillation features and the transient power distribution map are cross-fused in the time domain and frequency domain to separate the electromechanical coupling fatigue stress spectrum of the medical equipment under physical excitation in the powered operation state. S3. At the edge computing node, the rated loss cycle curve of the medical equipment is extracted from the equipment attribute library as the attenuation baseline. Adaptive nonlinear mapping stretching is performed on the attenuation baseline through the characteristic amplitude of each frequency band in the electromechanical coupling fatigue stress spectrum, and the dynamic health margin feature vector under the current circulation node of the medical equipment is obtained by parsing. S4. Receive real-time clinical diagnosis and treatment scenarios and operational precision access conditions for medical equipment issued by the clinical equipment resource scheduling platform. Perform multi-level degradation matching between the dynamic health margin feature vector and the operational precision access conditions to generate scenario access control instructions. Transmit the scenario access control instructions to the electronic ink RFID tag on the medical equipment side through the IoT edge gateway to trigger physical explicit locking of available clinical scenario permissions and terminal call authentication interception.

2. The method for dynamic sensing of medical equipment based on the Internet of Things according to claim 1, characterized in that: The specific process of synchronously acquiring the indoor positioning trajectory and corresponding multi-axis spatial oscillation sequence of medical equipment along the physical flow path across wards, and capturing the transient power distribution spectrum of the medical equipment power supply link under the same timestamp reference is as follows: By deploying indoor positioning beacons at medical equipment distribution nodes, radio frequency positioning tags on the chassis load-bearing structure of medical equipment are tracked in real time to generate indoor positioning trajectories with spatial coordinates. The IoT MEMS sensing terminal fixed at the interface between the chassis load-bearing structure and the core probe of the medical equipment is activated, and the three-dimensional acceleration vector sequence of the medical equipment during the propulsion process is extracted according to the preset sampling rate to construct a multi-axis spatial oscillation sequence. By connecting the Hall current sensor and voltage sampling module to the power supply bus of the medical equipment, the alternating current waveform and transient voltage drop amplitude of the power supply link of the medical equipment under energized operation are extracted and mapped to generate a transient power distribution map. The indoor positioning trajectory, multi-axis spatial oscillation sequence, and transient power distribution map are timestamped and encapsulated using the global unified clock pulse of the IoT edge gateway.

3. The method for dynamic sensing of medical equipment based on the Internet of Things according to claim 1, characterized in that: The specific process of spatially slicing the multi-axis spatial oscillation sequence along the indoor positioning trajectory of medical equipment and extracting the continuous envelope oscillation features within each spatial slice is as follows: Retrieve the indoor spatial layout map, map the indoor positioning trajectory onto the indoor spatial layout map, and mark the structural change coordinate nodes of the medical equipment flow across the height jump area; Extract the coordinate nodes of structural mutations, establish a dynamic spatial sliding window, and perform spatial dimension truncation and segmentation on the multi-axis spatial oscillation sequence through the dynamic spatial sliding window to generate independent spatial slices; For multi-axis spatial oscillation sequences within independent spatial slices, the instantaneous amplitude profile of the multi-axis spatial oscillation sequence is extracted by Hilbert transform, filtering out environmental noise components and preserving continuous envelope oscillation characteristics.

4. The method for dynamic sensing of medical equipment based on the Internet of Things according to claim 3, characterized in that: The specific process of separating the electromechanical coupling fatigue stress spectrum of medical equipment under physical excitation during energized operation by performing time-domain and frequency-domain cross-fusion of continuous envelope oscillation characteristics and transient power distribution spectrum is as follows: In the time domain, the peak moment of the physical shock wave in the continuous envelope oscillation features is extracted, and the current fluctuation range and power distortion range that occur synchronously with the peak moment of the physical shock wave in the transient power distribution spectrum are located. In the frequency domain, wavelet packet decomposition is performed simultaneously on the continuous envelope oscillation characteristics and power distortion interval to obtain the physical oscillation energy frequency band and the electrical interference frequency band of the same frequency resonance. A time-frequency cross-spectral density matrix is ​​constructed, and the cross-correlation characteristic matrix of the physical oscillation energy band and the electrical interference band is reconstructed. The mechanical vibration component and the power grid fluctuation component are separated to obtain the electromechanical coupling fatigue stress spectrum.

5. The method for dynamic sensing of medical equipment based on the Internet of Things according to claim 1, characterized in that: At the edge computing node, the specific process of extracting the rated wear cycle curve of medical equipment from the device attribute library as the attenuation baseline is as follows: Analyze the device physical medium access control address broadcast by the IoT MEMS sensing terminal, and build a low-level attribute addressing link for medical equipment type at the edge computing node; The device attribute library is retrieved along the underlying attribute addressing link to extract the set of calibration life thresholds and reference impedance feature profiles of electrical components of medical equipment; A polynomial time series fitting is performed on the set of calibrated lifetime thresholds of electrical components and the reference impedance characteristic profile to generate the rated loss cycle curve, and the rated loss cycle curve is loaded onto the edge computing node and configured as the attenuation baseline.

6. The method for dynamic sensing of medical equipment based on the Internet of Things according to claim 5, characterized in that: The specific process of obtaining the dynamic health margin feature vector of the medical equipment at the current circulation node by performing adaptive nonlinear mapping stretching on the attenuation baseline using the characteristic amplitudes of each frequency band in the electromechanical coupling fatigue stress spectrum is as follows: Perform band energy integration on the electromechanical coupling fatigue stress spectrum to extract the first and second characteristic amplitudes in the electromechanical coupling fatigue stress spectrum; A two-dimensional environmental stress penalty matrix is ​​constructed by using the first feature amplitude and the second feature amplitude. The two-dimensional environmental stress penalty matrix is ​​then subjected to tensor convolution with the current time node slope of the attenuation baseline to generate a nonlinear stretching mapping factor. Dynamic compression and amplitude distortion processing are performed on the time axis of the decay baseline by a nonlinear stretching mapping factor. The end state value of the baseline after dynamic compression and amplitude distortion processing is extracted, and the dynamic health margin feature vector is output.

7. The method for dynamic sensing of medical equipment based on the Internet of Things according to claim 1, characterized in that: The specific process of receiving real-time clinical diagnostic and treatment scenarios from the clinical equipment resource scheduling platform, and performing multi-level degradation matching between the dynamic health margin feature vector and the operational accuracy access conditions to generate scenario access control instructions is as follows: The received operational precision access conditions for medical equipment in different clinical diagnosis and treatment scenarios are analyzed into a standardized scenario constraint space that includes signal-to-noise ratio threshold, impedance tolerance boundary and immunity level. Project the dynamic health margin feature vector onto the standardized scene constraint space, and calculate the multidimensional spatial topological distance between the dynamic health margin feature vector and the boundary features of each level of clinical diagnosis and treatment scene in the standardized scene constraint space. Perform multi-level degradation matching based on the neighbor gradient of multi-dimensional spatial topological distance, remove mismatched scene labels, retain the whitelist of available scenes, and generate scene access control instructions.

8. The method for dynamic sensing of medical equipment based on the Internet of Things according to claim 7, characterized in that: The specific process of transmitting scene access control commands through the IoT edge gateway to the electronic ink RFID tag on the medical equipment side, triggering the physical explicit locking of available clinical scene permissions and the terminal call authentication and interception, is as follows: The communication protocol is encapsulated on the IoT edge gateway side to execute scene access control commands and send them to the microcontroller inside the electronic ink RFID tag on the medical equipment side; The microcontroller parses the available scenario whitelist in the scenario access control command, drives the display matrix of the electronic ink RFID tag to reconstruct the screen, presents the scenario graphic identifier, and performs physical explicit locking of the available clinical scenario permissions; The microcontroller synchronously overwrites the available scenario whitelist into the storage sector of the e-ink RFID tag, blocking unauthorized terminals from exceeding the handshake protocol of the storage sector and establishing a terminal call authentication interception state.

Images