Insulin pump flow monitoring method based on wireless networking

Abstract

The application discloses a kind of insulin pump flow monitoring methods based on wireless networking, it is related to wireless network communication field, including: constructing wireless monitoring networking based on multi-hop mesh topology;Insulin pump monitoring node is bound with insulin pump flow output end, sets flow acquisition frequency, data transmission period and flow standard threshold value, and configuration parameters are synchronized to relay node and terminal processing node by wireless link;The application selects and redundancy backup by optimizing wireless communication link, improves data transmission stability and coverage, adapts different scene signal environment, reduces delay and packet loss risk, guarantees flow data real-time and integrity;Dynamic adjustment acquisition frequency, fit medicine precision and flow fluctuation.

Description

Technical Field

[0001] This invention relates to the field of wireless network communication technology, specifically to a method for monitoring insulin pump flow based on wireless networking. Background Technology

[0002] Insulin pump flow monitoring is the core safety monitoring function of insulin pumps. It can collect flow-related data in real time during the infusion process, accurately monitor the basal infusion volume, the infusion rate of pre-meal bolus doses, and the cumulative infusion volume. It can quickly identify infusion problems such as tube blockage, flow interruption, and abnormal flow rate and provide immediate warnings, ensuring the accuracy and stability of insulin infusion and avoiding the risk of large fluctuations in blood glucose caused by infusion abnormalities.

[0003] Patent application No. 202410537242.8 discloses an LSTM-based method for predicting insulin pump infusion accuracy errors. This application aims to address the problem that "in the prior art, most insulin infusion devices are open-loop, rarely or not at all addressing precise drug delivery. Compared to MDI, CSII is significantly more effective in treating diabetes and has a lower risk of side effects. One of the key factors for the therapeutic effect of CSII is the accuracy of insulin pump delivery; inappropriate insulin dosage can lead to hyperglycemia, hypoglycemia, and even life-threatening situations in extreme cases. LSTM models have achieved many successful applications in time-series data research across various fields, including textual language modeling, speech recognition, machine translation, traffic flow prediction, and road transport-related applications. However, the application of LSTM models in predicting and compensating for insulin pump infusion accuracy errors has not yet been implemented."

[0004] However, based on the aforementioned publicly available patents, network communication security is also a crucial aspect of monitoring and maintenance for insulin pump flow monitoring scenarios.

[0005] To address this, we propose a wireless networking-based method for monitoring insulin pump flow. Summary of the Invention

[0006] In view of the above-mentioned shortcomings of the existing technology, the present invention provides an insulin pump flow monitoring method based on wireless networking, which can effectively solve the problems of the existing technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions;

[0008] This invention discloses a method for monitoring insulin pump flow based on wireless networking, comprising:

[0009] Construct a wireless monitoring network based on a multi-hop mesh topology;

[0010] Bind the insulin pump monitoring node to the insulin pump flow output terminal, set the flow acquisition frequency, data transmission cycle and flow standard threshold, and synchronize the configuration parameters to the relay node and terminal processing node via wireless link;

[0011] The insulin pump monitoring node collects insulin pump output flow data in real time, packages the data into data frames according to a preset format, and sends them to the relay node via wireless networking in a multi-hop transmission mode.

[0012] After receiving traffic data frames, the relay node verifies the integrity and rationality of the data, removes missing frames and abnormal data that exceed the reasonable range, and forwards the verified data frames to the terminal processing node according to the network link priority.

[0013] The terminal processing node performs real-time analysis and comparison of the verified flow data, matches the analysis results with the preset flow standard threshold, and immediately triggers an early warning signal when an abnormal flow is detected, and transmits the early warning information back to the insulin pump monitoring node and relay node.

[0014] The terminal processing node monitors the communication status of each node in real time. When transmission delay, signal attenuation or node abnormality is detected, the transmission power and communication channel of each node are dynamically adjusted.

[0015] The wireless monitoring network includes insulin pump monitoring nodes, relay nodes, and terminal processing nodes. Each node pre-complies with wireless communication protocol adaptation, establishes a two-way data transmission link, and provides conditions for collaborative communication between nodes.

[0016] Furthermore, the deployment of each node in the wireless monitoring network meets the requirements of coverage without blind spots and preset link redundancy backup;

[0017] Each node selects the optimal communication link through link quality assessment:

[0018] ;

[0019] In the formula: This represents the link reliability coefficient. , , These are signal strength weight, packet loss tolerance weight, and low latency weight, respectively. This is the normalized value of the inter-node communication signal strength; This refers to the packet loss rate of the link. This represents the actual transmission delay of the link. The preset maximum allowable transmission delay;

[0020] During network construction, each node should prioritize selecting Establish a communication connection on a link that is not lower than the preset link quality threshold.

[0021] Furthermore, the binding between the insulin pump monitoring node and the insulin pump flow output terminal is achieved through a dual binding mechanism of physical interface adaptation and wireless identifier matching: the physical interface adaptation is fixed to the insulin pump flow output port by mechanical engagement of the adapter connector, and the wireless identifier matching is achieved by associating and binding the unique device identifier of the monitoring node with the device serial number of the insulin pump and storing it in the device association library of the terminal processing node.

[0022] The flow collection frequency is set to conform to the following:

[0023] ;

[0024] In the formula: This refers to the real-time traffic collection frequency. This is the preset reference frequency, i.e., the initial sampling frequency; This is the frequency adjustment factor; The standard deviation of traffic within the most recent data transmission cycle; This is the current monitoring duration; The preset baseline monitoring cycle;

[0025] Among them, the real-time traffic collection frequency is subject to the upper and lower limits of the preset change threshold during the real-time dynamic change process based on the above formula.

[0026] Furthermore, the preset format data frame includes three parts: a frame header, a data segment, and a frame trailer. The frame header is 8 bytes and includes: monitoring node ID, data acquisition timestamp, and data type identifier; the data segment is 16 bytes and includes instantaneous traffic value, cumulative output traffic value, acquisition ambient temperature value, remaining battery power, and data status identifier; the frame trailer is 4 bytes and includes cyclic redundancy check code and link priority identifier.

[0027] After the insulin pump monitoring node collects flow data, it first filters the data, and then packages the preprocessed data into data frames according to a preset format.

[0028] ;

[0029] In the formula: This represents the filtered flow rate value collected in the kth sampling. This represents the filtered flow rate value collected in the (k-1)th sampling. Kalman gain; This represents the raw flow rate value collected in the kth sampling. Let be the covariance of the estimation error of the (k-1)th acquisition; To measure the noise covariance; Let $\mathbf{k}$ be the estimation error covariance of the $k$-th data collection. Let be the process noise covariance.

[0030] Furthermore, the data integrity verification in the relay node checks whether the frame header identifier, frame length, and frame tail check code of the data frame conform to the preset format. If any field is missing or incorrect, it is determined to be an invalid frame and discarded.

[0031] Data reasonableness verification is determined using a traffic reasonableness coefficient:

[0032] ;

[0033] In the formula: This is the traffic rationality coefficient; This is the instantaneous value of the traffic being verified. The reference flow rate value corresponding to the current dosing regimen of the insulin pump synchronized to the terminal processing node; This is a preset minimum constant; The standard deviation of the flow rate within the most recent preset time window;

[0034] The calculated Compare with the preset reasonableness threshold, if If the data is below a preset reasonableness threshold, it is considered abnormal data and is directly removed; if If the data is not lower than the preset reasonableness threshold, it is considered reasonable and the verification is considered passed.

[0035] The relay node sorts data frames that pass both integrity and rationality checks according to the link priority identifier, and prioritizes forwarding high-priority data frames. During the forwarding process, the data forwarding path and forwarding timestamp are recorded.

[0036] Furthermore, the real-time parsing of traffic data by the terminal processing node includes:

[0037] Extract the instantaneous flow rate, cumulative output flow rate, and ambient temperature from the data frame, combine them with the insulin pump's dosing protocol information, including single dose, dosing duration, and dosing interval, and calculate the actual dosing progress.

[0038] The comparison process calculates the flow deviation:

[0039] ;

[0040] In the formula: For flow deviation; , Instantaneous flow deviation weight and cumulative flow deviation weight; The real-time instantaneous flow rate value obtained through analysis; The preset instantaneous flow rate standard threshold; The cumulative output flow value obtained from the analysis; The preset cumulative traffic standard threshold for the corresponding time period;

[0041] when When the flow rate is not lower than the preset deviation threshold, it is judged as abnormal; when If the flow rate is below the preset deviation threshold, it is considered normal.

[0042] Furthermore, the warning signal is categorized by deviation. Divided into three levels:

[0043] If the deviation falls within the first preset range, a Level 1 warning is issued. If the deviation falls within the second preset range, a Level II warning is issued. If the deviation is not lower than the third preset deviation threshold, it is a level three warning, and the upper limit of the first preset deviation interval is less than the upper limit of the second preset deviation interval, and the upper limit of the second preset deviation interval is less than the third preset deviation threshold.

[0044] A Level 1 warning triggers an audible and visual alert. The warning information includes the type of anomaly and the current deviation. The type of anomaly includes instantaneous deviation and cumulative deviation.

[0045] In addition to audible and visual alerts, Level 2 warnings send push notifications to pre-bound mobile devices.

[0046] The Level 3 warning system, building upon the Level 2 warning system, sends a flow pause request signal to the insulin pump and simultaneously records the complete data frame and network status information at the time of the anomaly.

[0047] Furthermore, the terminal processing node monitors the communication status of each node using metrics including transmission delay, signal attenuation, and node connectivity, in order to construct a communication quality assessment index.

[0048] ;

[0049] In the formula: It is a communication quality assessment index; For transmission delay; The preset maximum allowable transmission delay; Signal attenuation; The preset maximum allowable signal attenuation; denoted as node connectivity; a, b, and c represent delay weight, decay weight, and connectivity weight, respectively.

[0050] when When the quality falls below the preset communication quality threshold, the communication status is deemed abnormal, triggering a dynamic adjustment process for transmission parameters.

[0051] Furthermore, the dynamic adjustment process for transmission parameters includes:

[0052] ;

[0053] In the formula: This is the amount of power adjustment. This is the power adjustment coefficient; The target communication quality index is preset; It is a communication quality assessment index; This represents the node's current transmission power. Rated transmission power of the node; This is the adjusted transmission power; The maximum and minimum transmission power are preset.

[0054] Furthermore, during the communication channel adjustment phase, the terminal processing node collects the interference intensity and channel capacity of each candidate channel in real time to determine the channel suitability.

[0055] ;

[0056] In the formula: Channel capacity weights; Let be the capacity of candidate channel j; Let be the maximum capacity of candidate channel j; Let be the interference intensity of candidate channel j; The maximum permissible interference intensity;

[0057] Finally, the candidate channel with the highest channel adaptability is selected as the new communication channel.

[0058] Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects:

[0059] This invention improves data transmission stability and coverage by optimizing wireless communication link selection and redundancy backup, adapting to different signal environments, reducing latency and packet loss risks, and ensuring the real-time and integrity of traffic data. It dynamically adjusts the acquisition frequency to match medication accuracy and traffic fluctuations, and combines data filtering and dual verification to effectively eliminate abnormal data and improve monitoring accuracy. Simultaneously, it provides tiered early warnings based on the degree of deviation, promptly reporting anomalies while balancing alerts and safety control, reducing medication risks. Furthermore, it optimizes transmission power and communication channels in real time to adapt to complex communication environment changes, ensuring continuous reliability of the monitoring link. This provides comprehensive, accurate, and stable support for insulin dosing monitoring, adapting to different patient medication needs and usage scenarios. Attached Figure Description

[0060] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0061] Figure 1 This is a flowchart illustrating a method for monitoring insulin pump flow based on wireless networking. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0063] The present invention will be further described below with reference to embodiments.

[0064] Example 1:

[0065] This embodiment provides a method for monitoring insulin pump flow based on wireless networking, such as... Figure 1 As shown, it includes:

[0066] Construct a wireless monitoring network based on a multi-hop mesh topology;

[0067] The deployment of each node in the wireless monitoring network meets the requirements of coverage without blind spots and preset link redundancy backup.

[0068] Each node selects the optimal communication link through link quality assessment:

[0069] ;

[0070] In the formula: This represents the link reliability coefficient. , , These are signal strength weight, packet loss tolerance weight, and low latency weight, respectively. This is the normalized value of the inter-node communication signal strength; This refers to the packet loss rate of the link. This represents the actual transmission delay of the link. The preset maximum allowable transmission delay;

[0071] The above formula integrates the three key influencing factors in wireless networking: signal strength, packet loss resistance, and transmission delay. By normalizing the signal strength, converting the packet loss rate and transmission delay indicators, and combining them with flexibly adjustable weight allocation, the weights can be dynamically adapted according to the differences in usage scenarios. For example, when there are many obstructions or large signal fluctuations, the signal strength is emphasized; when the risk of packet loss is high, the packet loss resistance weight is strengthened; and when timeliness requirements are high, the low latency weight is increased. This allows for the accurate selection of the optimal communication link and supports automatic switching to the suboptimal link in case of failure, thereby meeting the requirements of network coverage without blind spots and redundancy backup.

[0072] During network construction, each node should prioritize selecting Establish communication connections on links that are not lower than the preset link quality threshold, and automatically switch to the second-best link when the best link fails.

[0073] in, , , The values ​​of all values ​​are in the range (0,1), and , , The sum is 1. When the insulin pump is used in an environment with many obstructions or where the wireless signal is easily interfered with, causing large fluctuations in signal strength, A larger value indicates that the signal is more likely to be effective when the application scenario is open and unobstructed, the signal transmission environment is stable, and the signal strength fluctuation is small. The smaller the value, the better; this applies when insulin pump flow monitoring has extremely high requirements for data integrity (e.g., accurate dosing depends on complete flow data) and when the link transmission is susceptible to external interference leading to a high risk of packet loss. A larger value indicates that when the link transmission environment is stable and the packet loss rate is within the preset low-risk range, the situation is more favorable. The smaller the value, the better; this is especially important when there are abnormal insulin pump flow rates requiring rapid alerts (such as medication monitoring in patients with acute diabetes) or when data transmission timeliness is critical. The larger the value, the lower the sensitivity of the monitoring scenario to transmission delay, and the more data can be transmitted within a preset delay range. The smaller the value;

[0074] Bind the insulin pump monitoring node to the insulin pump flow output terminal, set the flow acquisition frequency, data transmission cycle and flow standard threshold, and synchronize the configuration parameters to the relay node and terminal processing node via wireless link;

[0075] The binding between the insulin pump monitoring node and the insulin pump flow output terminal is achieved through a dual binding mechanism of physical interface adapter and wireless identifier matching: the physical interface adapter is mechanically fixed to the insulin pump flow output port through the adapter connector, and the wireless identifier matching is achieved by associating and binding the unique device identifier of the monitoring node with the device serial number of the insulin pump and storing it in the device association library of the terminal processing node;

[0076] The traffic sampling frequency is set according to the following:

[0077] ;

[0078] In the formula: This refers to the real-time traffic collection frequency. This is the preset reference frequency, i.e., the initial sampling frequency; This is the frequency adjustment factor; The standard deviation of traffic within the most recent data transmission cycle; This is the current monitoring duration; The preset baseline monitoring cycle;

[0079] The above formula is based on a preset reference frequency. It takes into account the flow fluctuation in the most recent transmission cycle and the current monitoring duration. By adjusting the coefficient, the acquisition frequency is directly related to the insulin pump's drug delivery accuracy requirements and the impact of flow fluctuation on medication safety. It can dynamically optimize the acquisition frequency as flow fluctuation intensifies or the monitoring process progresses, ensuring that enough key data is obtained. At the same time, it avoids the frequency being too high or too low by setting upper and lower limits, thus ensuring the monitoring effect while taking into account the efficiency of resource utilization.

[0080] Among them, the real-time traffic collection frequency is subject to the upper and lower limits of the preset change threshold during the real-time dynamic change process based on the above formula;

[0081] The value of ∈[0,1] is positively correlated with the accuracy requirements of insulin pump administration and the degree of influence of flow fluctuation on medication safety;

[0082] The insulin pump monitoring node collects insulin pump output flow data in real time, packages the data into data frames according to a preset format, and sends them to the relay node via wireless networking in a multi-hop transmission mode.

[0083] The preset format data frame consists of three parts: frame header, data segment, and frame trailer. The frame header is 8 bytes, including: monitoring node ID (3 bytes), data acquisition timestamp (4 bytes), and data type identifier (1 byte). The data segment is 16 bytes, including instantaneous flow value (4 bytes), cumulative output flow value (6 bytes), acquisition ambient temperature value (2 bytes), remaining battery power (2 bytes), and data status identifier (2 bytes). The frame trailer is 4 bytes, including cyclic redundancy check code (2 bytes) and link priority identifier (2 bytes).

[0084] After the insulin pump monitoring node collects flow data, it first filters the data, and then packages the preprocessed data into data frames according to a preset format.

[0085] ;

[0086] In the formula: This represents the filtered flow rate value collected in the kth sampling. This represents the filtered flow rate value collected in the (k-1)th sampling. Kalman gain; This represents the raw flow rate value collected in the kth sampling. Let be the covariance of the estimation error of the (k-1)th acquisition; To measure the noise covariance; Let $\mathbf{k}$ be the estimation error covariance of the $k$-th data collection. For process noise covariance;

[0087] The above formula dynamically calculates the Kalman gain by associating the previous filtering result, the estimation error covariance, and the original acquisition data. The gain is flexibly adjusted according to the estimation error at the previous moment and the inherent measurement error of the acquisition device. The initial parameters during the first filtering are calibrated based on historical flow measurement error data of similar insulin pumps. Subsequently, the dynamic fluctuations of the insulin pump flow output are continuously adapted through iterative formulas, effectively filtering out noise interference in the original data and improving the accuracy and stability of the flow data.

[0088] Where, when k=1, The initial raw flow rate of the insulin pump output collected by the insulin pump monitoring node. The value is the preset initial flow rate corresponding to this insulin pump delivery protocol. The value is set to the preset initial estimation error covariance;

[0089] ∈[0,1], the larger the estimation error covariance of the previous moment and the smaller the inherent measurement error of the flow acquisition device, the larger the value, and vice versa; the estimation error covariance is initially set based on the preset initial value calibrated by the historical flow measurement error data of similar insulin pumps, and the subsequent value is calculated by combining the estimation error covariance of the previous moment, Kalman gain and dynamic fluctuation error of insulin pump flow output through the filtering iteration formula.

[0090] After receiving traffic data frames, the relay node verifies the integrity and rationality of the data, removes missing frames and abnormal data that exceed the reasonable range, and forwards the verified data frames to the terminal processing node according to the network link priority.

[0091] Data integrity verification in relay nodes checks whether the frame header identifier, frame length, and frame tail check code of the data frame conform to the preset format. If any field is missing or incorrect, the frame is determined to be invalid and discarded.

[0092] Data reasonableness verification is determined using a traffic reasonableness coefficient:

[0093] ;

[0094] In the formula: This is the traffic rationality coefficient; This is the instantaneous value of the traffic being verified. The reference flow rate value corresponding to the current dosing regimen of the insulin pump synchronized to the terminal processing node; This is a preset minimum constant; The standard deviation of the flow rate within the most recent preset time window;

[0095] The above formula starts from two core dimensions: the deviation between the current instantaneous flow rate and the reference flow rate, and the flow rate fluctuation within the most recent preset time window. It introduces a minimum constant to avoid the abnormal situation of the denominator being zero during the calculation process. By comprehensively quantifying the indicators of these two dimensions, a rationality coefficient is formed. This can accurately identify instantaneous abnormal data that exceeds the normal dosing range, and also eliminate unreliable data with excessive short-term flow rate fluctuations, thereby fully ensuring that the flow rate data transmitted to the terminal conforms to the actual medication scenario.

[0096] The calculated Compare with the preset reasonableness threshold, if If the data is below a preset reasonableness threshold, it is considered abnormal data and is directly removed; if If the data is not lower than the preset reasonableness threshold, it is considered reasonable and the verification is considered passed.

[0097] Relay nodes sort data frames that pass both integrity and rationality checks according to link priority identifiers, and prioritize forwarding higher priority data frames. During the forwarding process, the data forwarding path and forwarding timestamp are recorded.

[0098] in, The range of values ​​is ;

[0099] The terminal processing node performs real-time analysis and comparison of the verified flow data, matches the analysis results with the preset flow standard threshold, and immediately triggers an early warning signal when an abnormal flow is detected, and transmits the early warning information back to the insulin pump monitoring node and relay node.

[0100] The real-time parsing of traffic data by the terminal processing node includes:

[0101] Extract the instantaneous flow rate, cumulative output flow rate, and ambient temperature from the data frame, combine them with the insulin pump's dosing protocol information, including single dose, dosing duration, and dosing interval, and calculate the actual dosing progress.

[0102] The comparison process calculates the flow deviation:

[0103] ;

[0104] In the formula: For flow deviation; , Instantaneous flow deviation weight and cumulative flow deviation weight; The real-time instantaneous flow rate value obtained through analysis; The preset instantaneous flow rate standard threshold; The cumulative output flow value obtained from the analysis; The preset cumulative traffic standard threshold for the corresponding time period;

[0105] The above formula takes into account the deviation between real-time instantaneous flow and cumulative output flow, and reflects the differentiated needs of different drug administration stages through weight allocation. Overall, it comprehensively and accurately reflects the degree of deviation between the actual flow and the standard threshold, providing support for subsequent anomaly judgment.

[0106] when When the flow rate is not lower than the preset deviation threshold, it is judged as abnormal; when When the flow rate is below the preset deviation threshold, it is considered normal.

[0107] in, , The values ​​of both are within the range of (0,1), and the sum of the two is 1. The value of is directly proportional to the real-time flow control priority of the insulin pump during the current administration phase. The value of is directly proportional to the accuracy requirement of the cumulative flow rate within the preset dosing cycle of the insulin pump;

[0108] Warning signals according to deviation Divided into three levels:

[0109] If the deviation falls within the first preset range, a Level 1 warning is issued. If the deviation falls within the second preset range, a Level II warning is issued. If the deviation is not lower than the third preset deviation threshold, it is a level three warning, and the upper limit of the first preset deviation interval is less than the upper limit of the second preset deviation interval, and the upper limit of the second preset deviation interval is less than the third preset deviation threshold.

[0110] A Level 1 warning triggers an audible and visual alert. The warning information includes the type of anomaly and the current deviation. The type of anomaly includes instantaneous deviation and cumulative deviation.

[0111] In addition to audible and visual alerts, Level 2 warnings send push notifications to pre-bound mobile devices.

[0112] The Level 3 warning system, based on the Level 2 warning system, sends a flow pause request signal to the insulin pump and simultaneously records the complete data frame and network status information when the anomaly occurs.

[0113] The operation of sending a flow pause request signal to the insulin pump must be performed only after the insulin pump itself has performed a safety check and confirmed it.

[0114] The terminal processing node monitors the communication status of each node in real time. When transmission delay, signal attenuation or node abnormality is detected, the transmission power and communication channel of each node are dynamically adjusted.

[0115] The terminal processing node monitors the communication status of each node using metrics including transmission delay, signal attenuation, and node connectivity, in order to construct a communication quality assessment index.

[0116] ;

[0117] In the formula: It is a communication quality assessment index; For transmission delay; The preset maximum allowable transmission delay; Signal attenuation; The preset maximum allowable signal attenuation; denoted as node connectivity; a, b, and c represent delay weight, decay weight, and connectivity weight, respectively.

[0118] The above formula integrates transmission delay, signal attenuation and node connectivity. Each indicator is transformed into a normalized evaluation factor by comparing it with the preset maximum value. At the same time, it allows users to customize the weight of the three indicators according to their own needs, so as to comprehensively cover the key factors affecting the communication status and adapt to the different users' different focus on delay, signal strength and connectivity stability, thereby accurately determining whether the communication status is normal.

[0119] when When the quality falls below the preset communication quality threshold, it is determined to be an abnormal communication status, triggering a dynamic adjustment process for transmission parameters;

[0120] The values ​​of a, b, and c are defined by the user, and a, b, and c are all positive numbers, and the sum of a, b, and c is 1.

[0121] The dynamic adjustment process for transmission parameters includes:

[0122] ;

[0123] In the formula: This is the amount of power adjustment. This is the power adjustment coefficient; The target communication quality index is preset; It is a communication quality assessment index; This represents the node's current transmission power. Rated transmission power of the node; This is the adjusted transmission power; The maximum and minimum transmission power are preset.

[0124] The above formula takes the deviation between the current communication quality and the preset target index as the core basis, and calculates the adjustment amount by combining the node's current transmission power and rated power. The adjustment coefficient will change dynamically according to the size of the deviation. The larger the deviation, the stronger the adjustment. At the same time, the adjusted power is strictly constrained within the preset maximum and minimum range. This can not only respond quickly to the decline in communication quality and improve the transmission effect through power optimization, but also avoid excessive power causing resource waste or excessive power affecting communication.

[0125] in, ∈[0,1], the larger the difference between the communication quality assessment index and the preset target communication quality index, the larger the value; the smaller the difference, the smaller the value.

[0126] During the communication channel adjustment phase, the terminal processing node collects the interference intensity and channel capacity of each candidate channel in real time to determine the channel suitability.

[0127] ;

[0128] In the formula: Channel capacity weights; Let be the capacity of candidate channel j; Let be the maximum capacity of candidate channel j; Let be the interference intensity of candidate channel j; The maximum permissible interference intensity;

[0129] The above formula takes into account both the transmission capacity and anti-interference capability of the candidate channel. By balancing the importance of the two in communication through weighting, the interference intensity can be flexibly quantified by the difference between the average power of the non-target signal and the channel noise floor power, or the normalized value of the interference signal power. It can accurately select the optimal channel with strong transmission capability and low interference to adapt to the complex channel environment in wireless networking and ensure the efficiency and stability of data transmission.

[0130] Finally, the candidate channel with the highest channel adaptability is selected as the new communication channel.

[0131] in, ∈ (0,1), Quantization is achieved by monitoring the difference between the average power of the non-target communication signal and the channel noise floor power, or by using the normalized value of the interference signal power relative to the maximum permissible interference power of the channel.

[0132] The wireless monitoring network includes insulin pump monitoring nodes, relay nodes, and terminal processing nodes. Each node pre-complies with wireless communication protocol adaptation, establishes a two-way data transmission link, and provides conditions for collaborative communication between nodes.

[0133] The method described in the above embodiments can capture traffic data in real time and accurately, quickly identify anomalies and issue timely warnings, effectively ensuring medication safety. It is also adaptable to different usage scenarios, has strong anti-interference capabilities, low transmission latency, stable and reliable data transmission, and can dynamically optimize communication status to reduce the impact of various anomalies, making monitoring continuous and efficient.

[0134] Application example:

[0135] A type 2 diabetes patient routinely administers insulin via a home pump. A wireless network-based traffic monitoring method is employed to ensure medication safety. First, a multi-hop mesh topology wireless monitoring network is established, consisting of one insulin pump monitoring node, two relay nodes, and one terminal processing node (a home smart terminal). During deployment, coverage is ensured across the patient's bedroom, living room, kitchen, and other activity areas without blind spots, and link redundancy is met. Each node establishes a connection by evaluating and selecting the optimal communication link.

[0136] The insulin pump monitoring node is dually bound to the insulin pump flow output terminal: it is mechanically fixed to the pump body's flow output port via an adapter connector, and the unique device identifier of the monitoring node is associated with the insulin pump serial number and stored in the device association library of the terminal processing node. A preset baseline flow rate acquisition frequency of 10 times / minute and a data transmission cycle of 5 minutes are set. A standard flow threshold is set based on the patient's dosing regimen (3 units per dose, 1-hour dosing duration, and 8-hour dosing interval). These configuration parameters are then synchronized to the relay node and the terminal processing node via a wireless link.

[0137] The monitoring node begins collecting insulin pump output flow data in real time according to the settings. The raw instantaneous flow rate of 5.2 units / hour is processed by Kalman filtering to obtain a filtered instantaneous flow rate of 5.1 units / hour. Then, data frames are packaged according to a preset format, including the monitoring node ID, data acquisition timestamp, instantaneous flow rate of 5.1 units / hour, cumulative output flow rate of 2.3 units, ambient temperature of 25℃, and remaining battery power of 85%. This data is then transmitted to the relay node via a multi-hop wireless network.

[0138] After receiving a data frame, the relay node first checks the frame header identifier, frame length, and frame tail check code to confirm that they all conform to the preset format and completes the data integrity verification. Then, it calculates the traffic rationality coefficient as 0.92, which is higher than the preset rationality threshold of 0.7. The data is then judged to be reasonable. After the verification is passed, the data is forwarded to the terminal processing node according to the link priority identifier, and the forwarding path and timestamp are recorded synchronously.

[0139] The terminal processing node analyzes the received data in real time, calculates the actual drug administration progress based on the patient's dosing schedule, and further calculates the flow deviation as 0.08, which is lower than the preset deviation threshold of 0.15, thus determining the flow to be normal. During monitoring, the terminal processing node continuously tracks the communication status of each node, calculating a communication quality assessment index of 0.85, which is lower than the preset target communication quality index of 0.95. It then dynamically adjusts the transmission power of the monitoring node from 15dBm to 17dBm. Simultaneously, it collects the interference intensity and channel capacity of each candidate channel, calculating that channel 3 has the highest suitability of 0.93, and switches the communication channel to channel 3 to ensure stable communication.

[0140] In a subsequent monitoring session, the flow rate data was processed and calculated to have a flow rate deviation of 0.25. The terminal processing node immediately triggered a level-two alert, activating an audio-visual prompt and sending a push notification to the patient's pre-bound mobile terminal, specifying the anomaly type as cumulative deviation and the current deviation level. If the flow rate deviation subsequently reaches 0.4, the terminal processing node will trigger a level-three alert. Based on the level-two alert operation, it will send a flow rate suspension request to the insulin pump. After confirmation by the insulin pump's own safety verification, the flow rate suspension operation will be executed. At the same time, the data frame and network status information at the time of the anomaly will be fully recorded to ensure patient medication safety.

[0141] In summary, the methods described in the above embodiments improve data transmission stability and coverage by optimizing wireless communication link selection and redundancy backup, adapting to different signal environments, reducing latency and packet loss risks, and ensuring the real-time performance and integrity of traffic data. They dynamically adjust the acquisition frequency to match medication accuracy and traffic fluctuations, and combine data filtering and dual verification to effectively eliminate abnormal data and improve monitoring accuracy. Furthermore, they provide tiered early warnings to promptly report anomalies based on the degree of deviation, balancing alerts and safety control to reduce medication risks. Real-time optimization of transmission power and communication channels adapts to complex communication environment changes, ensuring continuous reliability of the monitoring link. This provides comprehensive, accurate, and stable support for insulin dosing monitoring, adapting to different patient medication needs and usage scenarios.

[0142] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring insulin pump flow rate based on wireless networking, characterized in that, include: Construct a wireless monitoring network based on a multi-hop mesh topology; Bind the insulin pump monitoring node to the insulin pump flow output terminal, set the flow acquisition frequency, data transmission cycle and flow standard threshold, and synchronize the configuration parameters to the relay node and terminal processing node via wireless link; The insulin pump monitoring node collects insulin pump output flow data in real time, packages the data into data frames according to a preset format, and sends them to the relay node via wireless networking in a multi-hop transmission mode. After receiving traffic data frames, the relay node verifies the integrity and rationality of the data, removes missing frames and abnormal data that exceed the reasonable range, and forwards the verified data frames to the terminal processing node according to the network link priority. The terminal processing node performs real-time analysis and comparison of the verified flow data, matches the analysis results with the preset flow standard threshold, and immediately triggers an early warning signal when an abnormal flow is detected, and transmits the early warning information back to the insulin pump monitoring node and relay node. The terminal processing node monitors the communication status of each node in real time. When transmission delay, signal attenuation or node abnormality is detected, the transmission power and communication channel of each node are dynamically adjusted. The wireless monitoring network includes insulin pump monitoring nodes, relay nodes, and terminal processing nodes. Each node pre-complies with wireless communication protocol adaptation, establishes a two-way data transmission link, and provides conditions for collaborative communication between nodes. The binding between the insulin pump monitoring node and the insulin pump flow output terminal is achieved through a dual binding mechanism of physical interface adaptation and wireless identifier matching: physical interface adaptation is achieved by mechanically engaging and fixing the adapter connector to the insulin pump flow output port, and wireless identifier matching is achieved by associating and binding the unique device identifier of the monitoring node with the device serial number of the insulin pump and storing it in the device association library of the terminal processing node. The flow collection frequency is set to conform to the following: ； In the formula: This refers to the real-time traffic collection frequency. This is the preset reference frequency, i.e., the initial sampling frequency; This is the frequency adjustment factor; The standard deviation of traffic within the most recent data transmission cycle; This is the current monitoring duration; The preset baseline monitoring cycle; Among them, the real-time traffic collection frequency is subject to the upper and lower limits of the preset change threshold during the real-time dynamic change process based on the above formula; The data integrity verification in the relay node checks whether the frame header identifier, frame length and frame tail check code of the data frame conform to the preset format. If any field is missing or incorrect, it is determined to be an invalid frame and discarded. Data reasonableness verification is determined using a traffic reasonableness coefficient: ； In the formula: This is the traffic rationality coefficient; This is the instantaneous value of the traffic being verified. The reference flow rate value corresponding to the current dosing regimen of the insulin pump synchronized to the terminal processing node; This is a preset minimum constant; The standard deviation of the flow rate within the most recent preset time window; The calculated Compare with the preset reasonableness threshold, if If the data is below a preset reasonableness threshold, it is considered abnormal data and is directly removed; if If the data is not lower than the preset reasonableness threshold, it is considered reasonable and the verification is considered passed. The relay node sorts data frames that pass both integrity and rationality checks according to the link priority identifier, and prioritizes forwarding high-priority data frames. During the forwarding process, the data forwarding path and forwarding timestamp are recorded.

2. The method for monitoring insulin pump flow rate based on wireless networking according to claim 1, characterized in that, The deployment of each node in the wireless monitoring network meets the requirements of coverage without blind spots and preset link redundancy backup. Each node selects the optimal communication link through link quality assessment: ； In the formula: This represents the link reliability coefficient. , , These are signal strength weight, packet loss tolerance weight, and low latency weight, respectively. This is the normalized value of the inter-node communication signal strength; This refers to the packet loss rate of the link. This represents the actual transmission delay of the link. The preset maximum allowable transmission delay; During network construction, each node should prioritize selecting Establish a communication connection on a link that is not lower than the preset link quality threshold.

3. The insulin pump flow monitoring method based on wireless networking according to claim 1, characterized in that, The preset format data frame includes three parts: frame header, data segment, and frame trailer. The frame header is 8 bytes and includes: monitoring node ID, data acquisition timestamp, and data type identifier; the data segment is 16 bytes and includes instantaneous flow value, cumulative output flow value, acquisition ambient temperature value, remaining battery power, and data status identifier; the frame trailer is 4 bytes and includes cyclic redundancy check code and link priority identifier. After the insulin pump monitoring node collects flow data, it first filters the data, and then packages the preprocessed data into data frames according to a preset format. ； In the formula: This represents the filtered flow rate value collected in the kth sampling. This represents the filtered flow rate value collected in the (k-1)th sampling. Kalman gain; This represents the raw flow rate value collected in the kth sampling. Let be the covariance of the estimation error for the (k-1)th acquisition; To measure the noise covariance; Let $\mathbf{k}$ be the estimation error covariance of the $k$-th data collection. Let be the process noise covariance.

4. The method for monitoring insulin pump flow rate based on wireless networking according to claim 1, characterized in that, The real-time parsing of traffic data by the terminal processing node includes: Extract the instantaneous flow rate, cumulative output flow rate, and ambient temperature from the data frame, combine them with the insulin pump's dosing protocol information, including single dose, dosing duration, and dosing interval, and calculate the actual dosing progress. The comparison process calculates the flow deviation: ； In the formula: For flow deviation; , Instantaneous flow deviation weight and cumulative flow deviation weight; The real-time instantaneous flow rate value obtained through analysis; The preset instantaneous flow rate standard threshold; The cumulative output flow value obtained from the analysis; The preset cumulative traffic standard threshold for the corresponding time period; when When the flow rate is not lower than the preset deviation threshold, it is judged as abnormal; when If the flow rate is below the preset deviation threshold, it is considered normal.

5. The insulin pump flow monitoring method based on wireless networking according to claim 4, characterized in that, The warning signal is based on the deviation. Divided into three levels: If the deviation falls within the first preset range, a Level 1 warning is issued. If the deviation falls within the second preset range, a Level II warning is issued. If the deviation is not lower than the third preset deviation threshold, it is a level three warning, and the upper limit of the first preset deviation interval is less than the upper limit of the second preset deviation interval, and the upper limit of the second preset deviation interval is less than the third preset deviation threshold; A Level 1 warning triggers an audible and visual alert. The warning information includes the type of anomaly and the current deviation. The type of anomaly includes instantaneous deviation and cumulative deviation. In addition to audible and visual alerts, Level 2 warnings send push notifications to pre-bound mobile devices. The Level 3 warning system, building upon the Level 2 warning system, sends a flow pause request signal to the insulin pump and simultaneously records the complete data frame and network status information at the time of the anomaly.

6. The method for monitoring insulin pump flow rate based on wireless networking according to claim 1, characterized in that, The terminal processing node monitors the communication status of each node using metrics including transmission delay, signal attenuation, and node connectivity, in order to construct a communication quality assessment index. ； In the formula: It is a communication quality assessment index; For transmission delay; The preset maximum allowable transmission delay; Signal attenuation; The preset maximum allowable signal attenuation; denoted as node connectivity; a, b, and c represent delay weight, decay weight, and connectivity weight, respectively. when When the quality falls below the preset communication quality threshold, the communication status is deemed abnormal, triggering a dynamic adjustment process for transmission parameters.

7. The insulin pump flow monitoring method based on wireless networking according to claim 6, characterized in that, The dynamic adjustment process for transmission parameters includes: ； In the formula: This is the amount of power adjustment. This is the power adjustment coefficient; The target communication quality index is preset; It is a communication quality assessment index; This represents the node's current transmission power. Rated transmission power of the node; This is the adjusted transmission power; The maximum and minimum transmission power are preset.

8. The method for monitoring insulin pump flow rate based on wireless networking according to claim 6, characterized in that, During the communication channel adjustment phase, the terminal processing node collects the interference intensity and channel capacity of each candidate channel in real time to determine the channel suitability. ； In the formula: Channel capacity weights; Let be the capacity of candidate channel j; Let be the maximum capacity of candidate channel j; Let be the interference intensity of candidate channel j; The maximum permissible interference intensity; Finally, the candidate channel with the highest channel adaptability is selected as the new communication channel.

Images