Field exercise and training medical service equipment management system based on block chain

Abstract

The invention discloses a field exercise and training medical service equipment management system based on a block chain, and relates to the technical field of field emergency medical equipment management and power supply safeguard.The viability of field medical service equipment in an extreme environment is remarkably improved through deep coupling of a block chain architecture and an intelligent algorithm; the dynamic analysis unit adopts an online sequence learning model to compress and analyze the window size according to the real-time altitude and power supply quality index, so that the capture speed of a plateau voltage sag event is increased to millisecond level, and the downtime risk caused by instantaneous power failure of key equipment such as a cardio-pulmonary resuscitation machine is thoroughly avoided; on the premise of protecting data privacy, the federal aggregation mechanism fuses fluctuation characteristics of a plurality of generators, the generated prediction model can adapt to different altitude intervals, and the problem of cross-scene failure of a single model is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of field emergency medical equipment management and power supply guarantee, and particularly relates to a field training medical service equipment management system based on a block chain. BACKGROUND

[0002] In field training medical service support in a high-altitude hypoxic area, a mobile medical team relies on a generator to supply power to key equipment such as a cardiopulmonary resuscitation machine; the air is thin under the plateau environment, which causes the combustion efficiency of the generator to be unstable, and superimposed sudden strong wind or low-temperature load changes, so that the output voltage of the power supply is prone to millisecond-level transient fluctuations; although the existing equipment management system is equipped with an intelligent voltage stabilizing module, the response mechanism thereof relies on fixed threshold judgment and cannot distinguish between the plateau-specific transient voltage drop and the conventional fluctuation; when the cardiopulmonary resuscitation machine encounters such a drop, the internal control circuit of the equipment may trigger a protective shutdown, causing the interruption of the first-aid process.

[0003] Some management schemes in the prior art use an LSTM time series prediction algorithm to give an early warning of power fluctuations, and voltage data is collected by a sensor deployed at the output end of the generator; the system establishes a fluctuation mode library and switches to a standby power supply in advance when a similar waveform is detected; however, the voltage fluctuations in the plateau environment have strong randomness, and the fluctuation characteristics are significantly different in different altitude intervals; the existing model needs to be retrained in a cross-altitude scene, and there is an identification blind area for transient drops with a duration of less than 100 ms; some systems attempt to increase the sampling frequency to capture transient characteristics, but a high-frequency data stream aggravates the processing delay of edge computing devices.

[0004] The limitations of such power management schemes are further transmitted to the medical service equipment coordination layer; when a single medical device is shut down due to voltage problems, the existing system cannot quickly deploy a nearby standby power supply; the resource scheduling logic thereof relies on a pre-set static priority list and cannot dynamically adjust the power supply strategy according to real-time first-aid needs; for example, when a cardiopulmonary resuscitation machine and a defibrillator are powered at the same time, the system will preferentially guarantee the device with higher power rather than the device with higher first-aid criticality; such a rigid management method will prolong the functional recovery time of the first-aid equipment in a plateau multi-device coordination scene. SUMMARY

[0005] In view of the above-mentioned existing problems, the present application is proposed.

[0006] The present application provides a field training medical service equipment management system based on a block chain, which solves the problem of medical device shutdown caused by millisecond-level drops in plateau power, poor cross-altitude adaptability of existing prediction models, and insufficient offline coordination capability.

[0007] To solve the above technical problems, the present application provides the following technical scheme:

[0008] The embodiment of the application provides a field training medical service equipment management system based on a blockchain, which comprises,

[0009] A blockchain network composed of a generator node, a medical device node and a command center node;

[0010] A voltage monitoring unit for collecting voltage time series data in real time at a generator output end and recording an altitude label for each data;

[0011] A dynamic analysis unit configured to process the time series data by using an online sequence learning model and dynamically adjust an analysis window size according to the altitude and a power supply quality index;

[0012] An intelligent contract unit for automatically issuing a standby power source switching instruction and broadcasting a state synchronization message containing a device identification and altitude information to associated medical devices when the dynamic analysis unit identifies a voltage sag event;

[0013] Wherein, the voltage sag event is determined according to a preset amplitude threshold and a duration threshold.

[0014] As a preferred scheme of the field training medical service equipment management system based on the blockchain, the dynamic analysis unit comprises a federal aggregation unit for aggregating fluctuation intensity indexes and load transition frequencies reported by multiple generator nodes without sharing original data, generating a prediction model suitable for different altitude intervals and issuing the prediction model to each node.

[0015] As a preferred scheme of the field training medical service equipment management system based on the blockchain, the blockchain network supports offline fault tolerance operation: a local node generates a message digest of the standby power source switching instruction, uses a node private key to digitally sign the digest, and binds the instruction plaintext and the signature as an offline data packet; when the network is restored, the signature is verified by using a corresponding public key, and the instruction and the verification result are synchronized to a blockchain ledger.

[0016] As a preferred scheme of the field training medical service equipment management system based on the blockchain, the voltage monitoring unit encapsulates the time series data into data blocks at a fixed time interval, and generates a chain storage record by attaching a device identification and an altitude label, wherein the on-chain storage contains an index and integrity check information, and the original data is saved in a local or edge storage.

[0017] As a preferred scheme of the field training medical service equipment management system based on the blockchain, the size compression of the analysis window satisfies at least one of the following trigger conditions:

[0018] a) The altitude is higher than a preset altitude threshold and the measured atmospheric pressure is lower than a preset air pressure threshold;

[0019] b) the generator output current has a high harmonic distortion;

[0020] The generator output current is sampled at a sampling period T s The generator output current sequence i(k) is collected, windowed and spectrum-estimated in a sliding window of length L to obtain the RMS of each harmonic component. To give the harmonic distortion, an index is constructed using the energy ratio of the fundamental and each harmonic to obtain the current harmonic distortion THD I (k) for mapping, the larger the quantity, the higher the harmonic content;

[0021]

[0022] where THD I (k) represents the current harmonic distortion at time k, dimensionless, I n,rms (k) represents the RMS current of the nth harmonic, unit A, subscript n represents the harmonic order, integer n≥2, represents the set of harmonic orders participating in the calculation, I 1,rms (k) represents the RMS current of the fundamental, unit A, subscript rms represents the root mean square measure, k represents the discrete time index, dimensionless, N H represents the highest harmonic order, dimensionless;

[0023] When the harmonic distortion increases, the analysis window is monotonically segmented and mapped to a compression coefficient c THD (k):

[0024]

[0025] where c THD (k) represents the window compression coefficient based on the harmonic distortion, the value range [c min ,1] is dimensionless, θ THD represents the standard limit threshold of the harmonic distortion, the starting point of triggering compression, dimensionless, Δ THD represents the threshold bandwidth from the starting point to the complete compression interval, dimensionless, β represents the compression slope of the linear segment, dimensionless, c min represents the minimum compression coefficient, dimensionless;

[0026] In the formula:

[0027] L'(k)=max(L min ,round(L·c THD (k))),

[0028] where L'(k) represents the length of the compressed sliding window, unit sample, Lmin represents the allowed minimum window length in samples, L represents the reference window length before compression in samples, and round(·) represents a rounding operator.

[0029] As a preferred scheme of the field training medical service equipment management system based on the blockchain, the intelligent contract unit is internally provided with a power supply priority rule library, the power supply priority is determined based on the weighted results of clinical task criticality, equipment recovery time and power demand, and the command center node is allowed to adjust the weight in real time according to the task label.

[0030] As a preferred scheme of the field training medical service equipment management system based on the blockchain, when multiple devices to be protected exist at the same time, the intelligent contract unit preferentially ensures that the cardiopulmonary resuscitation machine meets the service level requirements of the minimum operating voltage and continuous power supply time, then distributes the defibrillator, and the remainder is used for non-emergency critical equipment.

[0031] As a preferred scheme of the field training medical service equipment management system based on the blockchain, the blockchain network adopts a hybrid architecture: a private chain composed of generator nodes is used to execute power supply control transactions, an alliance chain composed of medical device nodes and command center nodes is used to manage equipment state and instruction logs, and state consistency is realized through cross-chain anchoring or relay mechanism.

[0032] As a preferred scheme of the field training medical service equipment management system based on the blockchain, the linkage of the dynamic analysis unit and the intelligent contract unit meets a preset maximum response delay threshold, so that the total delay from detecting the voltage sag event to issuing the standby power source switching instruction does not exceed the threshold;

[0033] The amplitude threshold and duration threshold criteria of the voltage sag event are defined as follows:

[0034] The sampling period T s Collect the generator output voltage sequence v(k); do DC removal and band-limited filtering on v(k), and calculate the window root mean square voltage V rms (k) as the amplitude representation in a sliding window with a length of L; refer to the voltage V ref Select the rated RMS or the sliding baseline of the latest steady state segment; enter the subsequent out-of-boundary and duration judgment;

[0035] When the drop ratio of V rms (k) relative to V ref exceeds the set amplitude threshold , enter the sag state and start accumulating the duration T con (k), which is incremented by T sOnce it is restored above the threshold, the state exits and T con (k) is reset to zero;

[0036] The amplitude and duration two conditions are combined into a one-time event judgment:

[0037]

[0038] Wherein, epsilon(k) represents the voltage sag event indication at time k, the value is 0 or 1, I[·] represents the indication function, 1 is taken when the condition is true, otherwise 0 is taken, V rms (k) represents the window root mean square voltage, unit V, subscript rms represents the root mean square metric, V ref represents the reference RMS voltage, unit V, represents the relative amplitude threshold, drop ratio, The superscript th represents the threshold, T con (k) represents the cumulative time of continuous drop state, unit ms, tau th represents the duration threshold, unit ms, k represents the discrete time index, dimensionless, and the symbol ∧ represents the logical and;

[0039] The value is set to 0.15-0.30, tau th The value is set to 20-100 ms; in field implementation, tau th Is converted into the minimum number of continuous samples for counting, and Is converted into the equivalent absolute voltage lower limit for panel checking; the timing adopts the monotonic accumulation logic of entering and accumulating and exiting and resetting, and the window update is used to eliminate jitter.

[0040] As a preferred scheme of the field medical service equipment management system based on the blockchain, the security mechanism of the system comprises: at least one combination of a national secret or an elliptic curve hash algorithm and a digital signature algorithm is used for chain storage, symmetric key encryption is used for state synchronization messages, and irreversible coding and access authorization control are set for device identification.

[0041] The present application has the beneficial effects that: through the deep coupling of the blockchain architecture and the intelligent algorithm, the survival ability of the field medical service equipment in the extreme environment is significantly improved. The dynamic analysis unit adopts an online sequence learning model, the analysis window size is compressed according to the real-time altitude and power quality index, the capture speed of the plateau voltage sag event is improved to the millisecond level, and the risk of downtime of key equipment such as cardiopulmonary resuscitation machine caused by instantaneous power failure is completely avoided. The federal aggregation mechanism fuses the fluctuation characteristics of multiple generators under the premise of protecting data privacy, and the generated prediction model can adapt to different altitude intervals, solving the single model cross-scene failure problem.

[0042] The smart contract unit sets power supply priority based on clinical first aid logic. When multiple device power supply conflicts, the minimum operating voltage and continuous power supply time of the cardiopulmonary resuscitation machine are prioritized, and then the defibrillator is distributed. The blockchain hybrid architecture separates power control instructions (private chain) from state management (consortium chain), meeting the real-time control needs of the generator node and ensuring the cross-department traceability of medical equipment status. The offline signature mechanism binds the hash digest and digital signature, so that the power switching instruction can be safely stored in the offline environment, and the synchronization is automatically verified after the network is restored, breaking the coordination failure caused by communication interruption in the wild.

[0043] The uplink storage of voltage monitoring data combined with integrity verification provides an unalterable traceability basis for equipment failure. The security mechanism uses the national encryption algorithm to encrypt the communication link and device identification, preventing sensitive medical data from being leaked. The present application identifies voltage sag to the total delay of power switching within the set threshold by responding to the delay threshold constraint, thereby eliminating the possibility of first aid equipment function interruption from the root. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope of the present application.

[0045] Figure 1 The framework schematic diagram of the field training medical service equipment management system in the embodiments. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solutions and advantages of the present application more clear, the following will further illustrate the present application in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0047] All terms used in the present application (including technical and scientific terms) have meanings generally understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted to have meanings consistent with the context of the present specification, and should not be interpreted in an idealized or overly formal manner.

[0048] For example, the terms "first", "second", etc. used in the present application are only used to distinguish similar objects, to distinguish the first object from another object, and are not used to describe a specific order or sequence, nor can be understood as indicating or implying relative importance.

[0049] The present application proposes a field training medical service equipment management system based on blockchain, which combines Figure 1 As shown in the figure, the system comprises:

[0050] a blockchain network composed of a generator node, a medical device node, and a command center node;

[0051] a voltage monitoring unit configured to collect voltage time series data at the generator output in real time and record an altitude label for each data;

[0052] In this embodiment, the altitude label is environmental height information attached to the voltage sample at the same sampling time, which can be obtained by GNSS height calculation or barometer-temperature joint estimation, and is aligned with the device local clock and the acquisition card clock to form the same timeline. Further, similarly, the voltage sampling period is 0.5-2 ms by default, with an adjustable range of 0.2-5 ms, and the altitude update period is 100-500 ms by default, with an adjustable range of 50-1000 ms. The threshold setting is based on the balance between the detection sensitivity of 10-100 ms drop events and the calculation load. Optionally, further, if the GNSS is blocked, the barometer height is used first, and the last GNSS lock value is used as the baseline for drift correction. In this embodiment, optionally, when the altitude or voltage data is missing for more than 1 s, the last valid label is maintained and a missing label is marked in the data block, and a lightweight integrity alarm is triggered on the edge side. The label is automatically removed after recovery.

[0053] a dynamic analysis unit configured to process the time series data using an online sequence learning model and dynamically adjust the analysis window size according to the altitude and power quality indicators; the power quality indicators at least include current harmonic distortion rate and voltage fluctuation intensity;

[0054] In this embodiment, specifically, the online sequence learning model refers to a time series model that is incrementally updated in the order of arrival without stopping the data stream. The engineering implementation can use OS-ELM or equivalent recursive least squares variants. The input is voltage time series and environmental labels, and the output is fluctuation intensity estimation and event indicator. Further, for example, the reference length of the analysis window is 4-10 fundamental periods by default, with an adjustable range of 2-20 fundamental periods. The window update period is recalculated every 1-5 sampling periods by default, with an adjustable range of 1-20 sampling periods. The setting is based on the compromise between transient capture ability and estimation stability. Optionally, similarly, when the edge algorithm utilization rate exceeds 80% or the temperature rise is out of limit, the window adjustment step is automatically halved and the model hyperparameters are temporarily frozen to maintain stable event criterion output. In this embodiment, further, the power quality indicators at least include current harmonic distortion rate and voltage fluctuation intensity, which are obtained by frequency domain estimation and time domain statistics, respectively. The update and window adjustment are completed in the same scheduling period to avoid race conditions.

[0055] The smart contract unit automatically issues a backup power switching instruction when the dynamic analysis unit identifies a voltage sag event, and broadcasts a state synchronization message containing the device identification and altitude information to the associated medical devices;

[0056] The voltage sag event is determined according to a preset amplitude threshold and a duration threshold.

[0057] In this embodiment, specifically, the amplitude threshold is given as a relative drop ratio based on the reference RMS voltage, and the duration threshold is given as the minimum cumulative duration of continuous out-of-range. Further, optionally, the default levels of both are issued and chained by the command center according to low / medium / high sensitivity, and the field can switch the levels according to the task situation without restarting the system. Similarly in this embodiment, to avoid jitter near the threshold, 1-2 sampling periods of hysteresis can be set for entering and exiting the sag state, and if the rated value reported by the device is missing, the sliding baseline of the latest steady state segment is used as the reference.

[0058] In one embodiment, the dynamic analysis unit includes a federal aggregation unit for aggregating the fluctuation intensity indicators and load transition frequencies reported by multiple generator nodes without sharing raw data, generating a prediction model suitable for different altitude intervals, and issuing it to each node.

[0059] In one embodiment, the blockchain network supports offline fault tolerance operation: the local node generates a message digest for the backup power switching instruction, digitally signs the digest using the node private key, and binds the instruction plaintext and signature as an offline data packet; when the network is restored, the signature is verified by the corresponding public key and the instruction and verification result are synchronized to the blockchain ledger;

[0060] Further in this embodiment, the offline data packet includes instruction plaintext, message digest, signature, timestamp, and node identification. The timestamp comes from a local high-stability clock and allows a clock deviation of ±2s. Specifically, optionally, the validity period of the offline data packet is 5-15 minutes by default, with an adjustable range of 1-60 minutes. Expiration packets are discarded before being chained and recorded in the local audit log. Similarly, for example, to suppress replay risks, the node writes the digest of the packet to the processed set after verification, and repeated digests are directly rejected without affecting subsequent synchronization.

[0061] In one embodiment, the voltage monitoring unit encapsulates the time series data into data blocks at fixed time intervals, and generates a chain storage record with device identification and altitude label, where on-chain storage includes index and integrity check information, and raw data is saved locally or in edge storage.

[0062] In this embodiment, specifically, the data block packaging interval is 200-500 ms by default, the adjustable range is 100-2000 ms, the index field at least contains start and end time, device identification, altitude range and digest check value, the digest adopts irreversible hash to support post verification; further, optionally, to reduce the load on the chain, the upper limit of the single block size is 64-256 kB by default, and if it exceeds the upper limit, it is rolled in time sequence; in this embodiment, similarly, the retention period of local or edge storage is not less than 7 days by default, the adjustable range is 3-30 days, and when the available storage space is less than 10%, the earliest data is discarded and the discarded digest is recorded on the chain.

[0063] In one embodiment, the size compression of the analysis window satisfies at least one of the following trigger conditions:

[0064] a) The altitude is higher than the preset altitude threshold and the measured atmospheric pressure is lower than the preset atmospheric pressure threshold;

[0065] b) The harmonic distortion rate of the generator output current is higher than the preset standard limit value;

[0066] In the generator output current, the sampling period T s Collect the generator output current sequence i(k), and perform windowing and spectrum estimation in the sliding window with length L to obtain the RMS of each order component; to give the harmonic distortion rate, an index is constructed using the energy ratio of the fundamental wave and each order harmonic, and the current harmonic distortion rate THD I (k) for mapping, the larger the quantity, the higher the harmonic content;

[0067]

[0068] where THD I (k) represents the current harmonic distortion rate at time k, dimensionless, I n,rms (k) represents the RMS current of the nth harmonic, unit A, subscript n represents the harmonic order, integer n≥2, represents the harmonic order set participating in the calculation, such as {2,…,N H}, I 1,rms (k) represents the RMS current of the fundamental wave, unit A, subscript rms represents the root mean square metric, k represents the discrete time index, dimensionless, N H represents the highest harmonic order, dimensionless;

[0069] When the harmonic distortion rate increases, the analysis window is compressed into a proportional coefficient c THD (k) in a monotonic segmented mapping:

[0070]

[0071] where c THD(k) denotes the window compression coefficient based on the harmonic distortion rate, dimensionless, [c min ,1] dimensionless, θ THD denotes the standard limit threshold of the harmonic distortion rate, the starting point of triggering compression, dimensionless, Δ THD denotes the threshold bandwidth from the starting point to the complete compression interval, dimensionless, β denotes the compression slope of the linear segment, dimensionless, c min denotes the minimum compression coefficient, dimensionless;

[0072] In the formula:

[0073] L'(k) = max(L min , round(L·c THD (k))),

[0074] wherein L'(k) denotes the compressed sliding window length, in samples, L min denotes the minimum window length allowed, in samples, L denotes the reference window length before compression, in samples, and round(·) denotes the rounding operator; the parameter θ THD may take 0.05-0.08, Δ THD may take 0.03-0.07, c min may take 0.3-0.6, and L min takes the lower limit satisfying the frequency resolution (such as being able to distinguish the fundamental ± first-order sideband) ;

[0075] Specifically, the present step takes the RMS energy ratio of the fundamental and each harmonic to constitute the current harmonic distortion rate, and further establishes a monotonic mapping from the distortion rate to the analysis window compression coefficient; the mapping takes the standard limit as the compression starting point, and sets a linear transition band and a lower limit coefficient, so that the window length is adaptively shrunk with the deterioration of the power quality, improving the time response to fast disturbances; when the distortion recovers to below the limit, the window returns to the reference length, maintaining the stability of the spectrum and statistical estimation;

[0076] The mapping only depends on a small number of parameters, facilitating calibration on different environments and models; Δ THD controls the transition bandwidth to suppress frequent swings near the threshold, c min and L min jointly serve as the lower bound to prevent statistical instability caused by excessive compression; the obtained L'(k) will directly drive the dynamic analysis unit to update the window configuration in the same time sequence pipeline, and participate in the subsequent response delay and contract triggering processes together with the event stream of the amplitude + duration criterion;

[0077] In the embodiment, specifically, the RMS and harmonic-related quantities are obtained by windowed spectral estimation or equivalent discrete transform, the window type can use a flat window or a Hanning window, and the window length is jointly defined by the compression coefficient and the minimum window length; further, similarly, if the fundamental RMS is lower than the minimum measurable threshold in a window, the distortion rate calculation of the window is skipped and the last valid value is retained to avoid the distortion of the ratio caused by zero value; optionally, further, the spectral estimation can also be realized by an equivalent frequency selection algorithm to reduce the calculation amount, and the output interface is consistent with the subsequent mapping without changing the trigger link; in the embodiment, optionally, when the estimation confidence of I1 is lower than the preset lower limit, the window compression is no longer tightened and is maintained at the last stable level to ensure the statistical stability.

[0078] In one embodiment, the smart contract unit has a power supply priority rule library built-in, determines the power supply priority based on the weighted results of clinical task criticality, device recovery time and power demand, and allows the command center node to adjust the weights in real time according to the task label;

[0079] In the embodiment, specifically, the clinical task criticality is mapped to a numerical weight by a task level label, the device recovery time is given by a historical restart time or a manufacturer's specified value, and the power demand is provided by a device nameplate or real-time measurement, and the three are generated into an ordering score according to a weighting rule; further, for example, the weights are configured in descending order of criticality > recovery time > power by default, the minimum granularity of weight update is a single device, and the update period is 30-120s by default, which can take effect immediately in an emergency state; optionally, similarly, to prevent chain actions caused by frequent switching, the contract side sets a minimum maintenance time for priority changes of the same device, which is 30-90s by default.

[0080] In one embodiment, the smart contract unit, when there are multiple devices to be protected at the same time, prioritizes to ensure that the cardiopulmonary resuscitation machine meets the service level requirements of the minimum operating voltage and continuous power supply time, and then allocates to the defibrillator, and the remainder is used for non-emergency critical devices;

[0081] In the embodiment, specifically, the minimum operating voltage is expressed as a proportion of the lower limit of the rated voltage of the device, and the continuous power supply time is expressed as the minimum maintenance time of a single task, both of which are given by the equipment manual and expert rules and can be issued before the task; further, optionally, when the capacity of the standby power supply is insufficient to simultaneously meet the two types of emergency devices, the contract first meets the minimum continuous time of the cardiopulmonary resuscitation machine, and then performs time slice allocation according to the minimum preparation time of the defibrillator, and the requests that are not met enter the queue and are prioritized to be dequeued when the capacity is restored; similarly, further, if the device self-check reports an abnormality or the temperature is out of limit, the contract temporarily suspends the priority of the device and records the reason.

[0082] In one embodiment, the blockchain network adopts a hybrid architecture: a private chain composed of generator nodes is used to execute power supply control type transactions, and a consortium chain composed of medical device nodes and command center nodes is used to manage equipment state and instruction logs, and the two achieve state consistency through cross-chain anchoring or relaying mechanisms;

[0083] In this embodiment, specifically, the cross-chain anchoring period is 5-15s by default, with an adjustable range of 2-60s, and the anchoring content includes the latest block height, cross-chain digest, and timestamp, which are used for state reconciliation of the private chain and the consortium chain. Further, optionally, if cross-chain relaying is unavailable for more than two anchoring periods, the private chain state is marked as uncertain on the consortium chain and is limited to read-only queries, and after recovery, the anchoring records are supplemented at one time. Similarly, further, to reduce the impact of network spikes, cross-chain messages can introduce a ±10% random jitter in the sending time to disperse traffic.

[0084] In one embodiment, the linkage of the dynamic analysis unit and the smart contract unit meets the preset maximum response delay threshold, so that the total delay from detecting a voltage sag event to issuing a standby power switching instruction does not exceed the threshold. In this embodiment, specifically, the total delay is composed of three segments: detection calculation, contract decision, and instruction issuance, with default upper limits of 20ms, 10ms, and 20ms, respectively, and adjustable ranges of 5-50ms, 5-30ms, and 10-80ms, respectively. Further, similarly, the edge periodically reports the end-to-end delay distribution and stores it on the chain, and if it exceeds the threshold for three consecutive times, it triggers the linkage strategy of downsampling and priority convergence. Optionally, further, when the delivery interface is congested, causing a timeout, the local controller triggers a one-time fallback switching and marks the event as an abnormal bypass, and after the network recovers, the records are supplemented.

[0085] The amplitude threshold and duration threshold criteria for voltage sag events are defined as follows:

[0086] With a sampling period T s Collect the generator output voltage sequence v(k); do DC removal and band-limited filtering on v(k), and calculate the window root mean square voltage V rms (k) as the amplitude representation within a sliding window of length L; refer to the voltage V ref Select the rated RMS or the sliding baseline of the latest steady state segment; enter the subsequent out-of-boundary and duration judgment;

[0087] When the drop ratio of V rms (k) relative to V ref exceeds the set amplitude threshold , enter the sag state and start accumulating the duration T con (k), which increases by T s every sampling period, and once it recovers above the threshold, the state exits and Tcon (k) reset to zero;

[0088] Combining the two conditions of amplitude and duration into a one-time event determination:

[0089]

[0090] where ε(k) represents the voltage sag event indicator at time k, taking value 0 or 1, I[·] represents the indicator function, taking 1 if the condition is true, otherwise 0, V rms (k) represents the window root mean square voltage, unit V, subscript rms represents the root mean square metric, V ref represents the reference RMS voltage, unit V, represents the relative amplitude threshold, drop ratio, superscript th represents threshold, T con (k) represents the cumulative time of continuous drop state, unit ms, τ th represents the duration threshold, unit ms, k represents the discrete time index, dimensionless, symbol ∧ represents logical and;

[0091] value is set to 0.15-0.30, τ th value is set to 20-100 ms; in field implementation, τ th is converted into the minimum number of continuous samples for counting, and τ is converted into the equivalent absolute voltage lower limit for panel checking; the timing adopts monotonic accumulation logic of entering to accumulate and exiting to reset, which is matched with window update to eliminate jitter;

[0092] Specifically, this step establishes a joint criterion of relative drop ratio and duration with window RMS as the amplitude representative, forming a closed loop from collection, boundary identification to event output;

[0093] The relative caliber is compatible with different generator operating points, and both reference methods of sliding baseline and rated value can adapt to field fluctuations; the accumulation of duration adopts the incremental and reset mechanism driven by state, which has inhibitory effect on isolated spikes, while maintaining sensitivity to sustained drops; in terms of parameter configuration, the amplitude threshold and duration threshold give the range available for engineering, which can be further refined when the downstream time delay threshold and device priority are linked, and when load disturbance or altitude leads to intensified power supply quality fluctuations, the thresholds can be scene-specifically calibrated under the premise of maintaining the same formula structure;

[0094] The event indicator serves as a unified entrance for subsequent dynamic analysis and intelligent contract triggering, and can be directly used for statistical frequency and response link delay, and combined with the priority rule library to complete power switching decision, so as to maintain the continuous availability of medical devices in complex field environments;

[0095] In this embodiment, specifically, the event indication quantity is written in the edge buffer in time series form and is rolled up by 1-2s granularity into three statistical quantities of occurrence rate, average duration and 95 percentile response time delay, which are used for situation assessment of the command center; further, optionally, if the occurrence rate in the continuous time window exceeds the preset threshold or the 95 percentile response time delay approaches 80% of the threshold, the command center automatically issues a higher sensitivity criterion gear and a stricter priority weight to relieve the risk in advance; similarly, further, when the statistical module detects data clock drift or sample gap, the corresponding time period is marked as not participating in evaluation and the existing strategy remains unchanged.

[0096] In one embodiment, the security mechanism of the system includes: at least one combination of national secret or elliptic curve type hash algorithm and digital signature algorithm is used for upper chain storage, symmetric key encryption is used for state synchronization message, and irreversible coding and access authorization control are set for device identification;

[0097] In this embodiment, specifically, the key is generated and distributed by the hardware root trust of the command center, the key rotation default period is 30-90 days, the adjustable range is 7-180 days, and the old key is retained for short-term decryption permission during the rotation period to support smooth switching; further, similarly, the replay protection window of the state synchronization message is 60-120s by default, the adjustable range is 10-300s, and the out-of-window message is directly discarded and counted; optionally, further, when an abnormal increase in signature verification failure rate or device identification conflict is detected, the system enters the reduced weight mode, only allows the minimum necessary read operation and stops strategy change, and is restored after manual confirmation.

[0098] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

[0099] In addition, those skilled in the art can understand that although some embodiments herein include certain features included in other embodiments rather than other features, the combination of features of different embodiments means to be within the scope of the present application and form different embodiments. For example, all the above embodiments can be used in any combination. The information disclosed in the BACKGROUND section is only intended to deepen the understanding of the overall background of the present application, and should not be regarded as acknowledging or implying in any form that the information constitutes prior art known to those skilled in the art.

Claims

1. A field training medical service equipment management system based on a blockchain, characterized in that, Consisting of, a blockchain network composed of generator nodes, medical device nodes, and command center nodes; a voltage monitoring unit for collecting voltage time series data at the generator output in real time and recording an altitude label for each data; a dynamic analysis unit configured to process the time series data using an online sequence learning model and dynamically adjust the analysis window size based on the altitude and power quality indicators; an intelligent contract unit that automatically issues a backup power switching instruction when the dynamic analysis unit identifies a voltage sag event and broadcasts a status synchronization message containing the device identification and altitude information to the associated medical devices; wherein the voltage sag event is determined according to a preset amplitude threshold and duration threshold.

2. The field training medical service equipment management system based on a blockchain according to claim 1, wherein, The dynamic analysis unit includes a federated aggregation unit for aggregating fluctuation intensity indicators and load transition frequencies reported by multiple generator nodes without sharing raw data, generating prediction models suitable for different altitude intervals, and distributing them to each node.

3. The field training medical service equipment management system based on the blockchain according to any one of claims 1 and 2, characterized in that, The blockchain network supports offline fault tolerance operation: local nodes generate a message digest for the backup power switching instruction, digitally sign the digest using a node private key, and bind the instruction plaintext and signature as an offline data packet; when the network is restored, the signature is verified using the corresponding public key, and the instruction and verification results are synchronized to the blockchain ledger.

4. The field training medical service equipment management system based on a blockchain according to claim 1, wherein, The voltage monitoring unit encapsulates the time series data into data blocks at fixed time intervals, and generates on-chain storage by attaching device identification and altitude labels, where on-chain storage includes index and integrity check information, and raw data is saved locally or in edge storage.

5. The field training medical service equipment management system based on blockchain of claim 1, wherein, The size compression of the analysis window satisfies at least one of the following trigger conditions: a) the altitude is higher than the preset altitude threshold and the measured atmospheric pressure is lower than the preset atmospheric pressure threshold; b) the harmonic distortion rate of the generator output current is higher than the preset standard limit; The generator output current is sampled at a sampling period T s The generator output current sequence i(k) is collected, windowed and spectrum estimated in a sliding window of length L to obtain the RMS of each order component; to give the harmonic distortion rate, the energy ratio of the fundamental and each order harmonic is used to construct an index to obtain the current harmonic distortion rate THD in the window I (k) for mapping, the larger the quantity, the higher the harmonic content; where THD I (k) represents the current harmonic distortion at time k, dimensionless, I n,rms (k) represents the RMS current of the nth harmonic, unit A, subscript n represents the harmonic order, integer n≥2, represents the set of harmonic orders participating in the calculation, I 1,rms (k) represents the fundamental RMS current, unit A, subscript rms represents the root mean square measure, k represents the discrete time index, dimensionless, N H represents the highest harmonic order, dimensionless; When the harmonic distortion ratio is raised, the analysis window is compressed by a monotonic piecewise mapping into a scale factor c THD (k): where c THD (k) denotes the window compression coefficient based on the harmonic distortion rate, dimensionless, in the interval [c min , 1], θ THD denotes the standard limit threshold of the harmonic distortion rate, the starting point of triggering compression, dimensionless, Δ THD denotes the threshold bandwidth from the starting point to the fully compressed interval, dimensionless, β denotes the compression slope of the linear segment, dimensionless, c min denotes the minimum compression coefficient, dimensionless; where: where L'(k) denotes the compressed sliding window length in samples, L min denotes the allowed minimum window length in samples, L denotes the reference window length in samples before compression, and round(·) denotes the rounding operator.

6. The field training medical service equipment management system based on blockchain of claim 1, wherein, The intelligent contract unit has a power supply priority rule library built-in, determines the power supply priority based on the weighted results of clinical task criticality, device recovery time, and power demand, and allows the command center node to adjust the weights in real time according to the task label.

7. The field training medical service equipment management system based on blockchain of claim 6, wherein, When there are multiple devices to be protected at the same time, the intelligent contract unit prioritizes ensuring that the cardiopulmonary resuscitation machine meets the minimum operating voltage and continuous power supply time service level requirements, followed by the defibrillator, and the remaining amount is used for non-emergency critical equipment. 8.The field training medical service equipment management system based on blockchain of claim 1, wherein, The blockchain network uses a hybrid architecture: the private chain composed of generator nodes is used to execute power control transactions, the alliance chain composed of medical device nodes and command center nodes is used for equipment status and instruction log management, and both achieve state consistency through cross-chain anchoring or relay mechanism. 9.The field training medical service equipment management system based on blockchain of claim 1, wherein, The linkage of the dynamic analysis unit and the intelligent contract unit meets the preset maximum response delay threshold, so that the total delay from detecting a voltage sag event to issuing a backup power switching instruction does not exceed the threshold; The amplitude threshold and duration threshold criteria for the voltage sag event are defined as follows: at a sampling period T s Collecting the generator output voltage sequence v(k); DC removing and band limiting filtering are performed on v(k), and the window root mean square voltage V rms (k) is calculated in a sliding window with a length of L, which is taken as the amplitude representation; the reference voltage V ref Selecting the rated RMS or the sliding baseline of the latest steady state section; entering the subsequent out-of-limit and duration judgment; When V rms (k) exceeds a set threshold value ref (k) falls below a set threshold value , enter a brownout state and begin accumulating a duration T con (k), incrementing T s (k) by one sample period, exit the state and reset T con (k) to zero once above the threshold value. Combine amplitude and duration conditions into a one-time event determination: where ε(k) denotes the voltage sag event indicator at time k, taking values in {0, 1}, I[·] denotes the indicator function, taking value 1 if the condition is true, and 0 otherwise, V rms (k) denotes the windowed root mean square voltage, in V, with the subscript rms denoting the root mean square measure, V ref denotes the reference RMS voltage, in V, denotes the relative magnitude threshold, the drop ratio, with the superscript th denoting the threshold, T con (k) denotes the accumulated time in the drop state, in ms, τ th denotes the duration threshold, in ms, k denotes the discrete time index, dimensionless, and the symbol ∧ denotes the logical and; Value set to 0.15~0.30, τ th Value set to 20~100ms; in field implementation, τ th Convert to the minimum number of continuous samples for counting, τ Convert to the equivalent absolute voltage lower limit for panel checking; use monotonic accumulation logic with entry accumulation and exit reset for timing, and cooperate with window update to eliminate jitter.

10. The field training medical service equipment management system based on blockchain of claim 1, wherein, The security mechanism of the system comprises: at least one combination of a national secret or an elliptic curve type hash algorithm and a digital signature algorithm is used for upper chain storage, symmetric key encryption is used for state synchronization messages, and irreversible coding and access authorization control are set for device identification.

Images