A data transmission management method and system for non-invasive ventilator

Abstract

The present invention relates to the technical field of medical equipment and data transmission, and discloses a data transmission management method and system for a non-invasive ventilator, comprising: inputting current flow characteristic data into a preset first machine learning model to obtain future flow characteristic data in a future time period; inputting the future flow characteristic data and an initial MTU size into a preset second machine learning model to obtain an optimal MTU size; when in the future time period, transmitting data message fragments of data generated by the non-invasive ventilator according to the optimal MTU size; screening out U lost or erroneous data packets and data packet characteristic data generated, obtaining priority coefficients of U lost or erroneous data packets according to the data packet characteristic data, and retransmitting the lost or erroneous data packets according to the priority coefficient; the present invention is conducive to improving the packet loss rate or retransmission amount of non-invasive ventilator data during transmission.

Description

Technical Field

[0001] The present invention relates to the technical field of medical equipment and data transmission, and more specifically, to a data transmission management method and system for a non-invasive ventilator. Background Art

[0002] With the widespread application of non-invasive ventilators in the medical field and technological advances, their functions are constantly enriched and upgraded. Not only basic respiratory parameters are required to be monitored, but also multiple physiological parameters such as blood oxygen saturation, heart rate, and end-tidal carbon dioxide concentration are required to be monitored, resulting in a significant increase in the type and volume of transmitted data. These data need to be transmitted to the central monitoring system in real time so that medical staff can conduct remote monitoring and diagnosis; however, due to the increase in data volume and type, the demand for transmission bandwidth increases accordingly, which in turn causes network congestion and increased instability in communication quality, causing a large number of data packets to be lost or errors to occur, thereby triggering frequent retransmission requests; and frequently initiating retransmission requests will not only further increase transmission delays, but also lead to unnecessary waste of network resources, affecting overall transmission efficiency; therefore, it is necessary to provide a data transmission management method that improves the transmission efficiency of non-invasive ventilators.

[0003] At present, there is a lack of transmission management methods or systems for the increase in the volume of non-invasive ventilator data. Although there are some relevant technical documents, for example, the patent with authorization announcement number CM111669406B discloses a non-invasive ventilator data security transmission method and system, which ensures the safe and lossless transmission of data by constructing a secure two-way data transmission channel; however, research and practical application of the above methods and prior arts have found that the above methods and prior arts have at least the following defects:

[0004] (1) It is impossible to dynamically adjust the MTU (i.e., maximum transmission unit, which indicates the maximum size (in bytes) of a data packet transmitted on a network connection), which makes it impossible to improve the throughput of non-invasive ventilator data and the transmission efficiency of data packets when the network is in good condition, and it is impossible to reduce the packet loss rate or retransmission volume when the network fluctuates, making it difficult to effectively improve the stability of network transmission;

[0005] (2) It is impossible to calculate the priority coefficients for different types of lost or erroneous data packets after dynamically adjusting the MTU size, and thus it is impossible to ensure the timeliness of the transmission of respiratory parameter data in non-invasive ventilator data. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a data transmission management method and system for a non-invasive ventilator.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A data transmission management method for a non-invasive ventilator, the method comprising:

[0009] Obtain current traffic characteristic data from time QM to time Q, and input the current traffic characteristic data into a preset first machine learning model for predicting future traffic characteristic data to obtain future traffic characteristic data in a future time period [Q+M, Q+2M], where Q and M are integers greater than zero;

[0010] Obtaining an initial MTU size, and inputting the future traffic feature data and the initial MTU size into a preset second machine learning model for predicting an optimal MTU size to obtain the optimal MTU size;

[0011] When in the future time period [Q+M, Q+2M], the initial MTU size is adjusted to the optimal MTU size, and the data message fragments of the data generated by the non-invasive ventilator are transmitted according to the optimal MTU size;

[0012] When in the future time period [Q+M, Q+2M], the U lost or erroneous data packets and data packet characteristic data generated are screened out, priority coefficients of the U lost or erroneous data packets are obtained according to the data packet characteristic data, and the lost or erroneous data packets are retransmitted according to the priority coefficients, where U is an integer greater than zero.

[0013] Furthermore, the current traffic characteristic data includes initial bandwidth utilization, initial delay, initial packet loss rate, initial jitter, initial data transmission frequency and initial network congestion coefficient at each moment;

[0014] The generation logic of the initial network congestion coefficient is as follows:

[0015] Acquire network congestion characterization data, wherein the network congestion characterization data includes network traffic, network bandwidth, average data packet arrival rate, and data packet service rate from time QM to Q;

[0016] Inputting the network congestion characterization data into a pre-built congestion coefficient calculation model to obtain an initial network congestion coefficient;

[0017] The expression of the pre-built congestion coefficient calculation model is as follows:

[0018] ;

[0019] Where: is the initial network congestion coefficient; is the network traffic at the i-th moment; is the network bandwidth; is the average packet arrival rate at the i-th moment; is the data packet service rate at the i-th moment; is the length of the time interval.

[0020] Furthermore, the generation logic of the preset first machine learning model is as follows:

[0021] Acquire historical traffic feature training data, and divide the historical traffic feature training data into a traffic feature training set and a traffic feature test set, wherein the historical traffic feature training data includes current traffic feature data from time QM to time Q and future traffic feature data in the corresponding future time period [Q+M, Q+2M];

[0022] The future traffic characteristic data includes the future bandwidth utilization, future delay, future packet loss rate, future jitter, future data transmission frequency and future network congestion coefficient at each moment in the future time period [Q+M, Q+2M];

[0023] Construct a first regression network, use the current traffic feature data in the traffic feature training set as the first regression network input data, and use the future traffic feature data in the future time period [Q+M, Q+2M] in the traffic feature training set as the output data of the first regression network, train the first regression network, and obtain a trained first machine learning network;

[0024] The trained first machine learning network is model verified using the traffic feature test set, and a first machine learning network with a value less than or equal to a first preset test error threshold is output as the preset first machine learning model.

[0025] Furthermore, the generation logic of the preset second machine learning model is as follows:

[0026] Acquire historical MTU size training data, and divide the historical MTU size training data into an MTU size training set and an MTU size test set, wherein the historical MTU size training data includes future traffic feature data and its corresponding optimal MTU size;

[0027] Constructing a second regression network, using the future traffic feature data in the MTU size training set as input data of the second regression network, and using the optimal MTU size in the MTU size training set as output data of the second regression network, training the second regression network, and obtaining a trained second machine learning network;

[0028] The trained second machine learning network is model verified using the MTU size test set, and a second machine learning network with a value less than or equal to a preset second test error threshold is output as the preset second machine learning model.

[0029] Further, the optimal MTU size is obtained by optimizing using a preconfigured genetic algorithm;

[0030] The optimization logic of the preconfigured genetic algorithm is as follows:

[0031] a1: Initialize the population: randomly generate an original population, which contains G individuals, each of which represents an MTU size, and G is an integer greater than zero;

[0032] a2: Fitness evaluation: Under each individual, detect the data retransmission amount, packet loss rate and actual network congestion coefficient of the transmission network; input the data retransmission amount and packet loss rate into the pre-built fitness function to calculate the fitness of each individual;

[0033] a3: Selection: Use the roulette method to select two individuals with high fitness in the original population as the father and mother;

[0034] a4: Crossover: Perform crossover operation on the father and mother to produce new individuals;

[0035] a5: Mutation: Perform mutation operation on the new individuals to obtain S new individuals, combine the S new individuals into a new population, replace the original population with the new population, and return to step a2;

[0036] a6: Repeat steps a2 to a5 above until the fitness of the individuals in the original population or the new population is greater than or equal to the preset fitness threshold, or the number of iterations is greater than or equal to the preset maximum number of iterations threshold, and output the MTU size represented by the corresponding individual as the optimal MTU size.

[0037] Furthermore, the calculation formula of the pre-constructed fitness function is: ; Where: For fitness, is the amount of data retransmission, is the packet loss rate.

[0038] Further, obtaining priority coefficients of U lost or erroneous data packets includes:

[0039] b1: extracting data packet characteristic data of the jth lost or erroneous data packet, including the data type and number of retransmissions of the data packet characteristic data;

[0040] b2: Input the data type into a predefined importance function to obtain the importance value of the lost or erroneous packet. ;

[0041] The expression of the importance function is as follows:

[0042] ;

[0043] Where: Indicates data type; Indicates that the data type is non-respiratory parameter data;

[0044] b3: Input the importance value and the number of retransmissions into the pre-built priority calculation model to obtain the priority coefficient of the jth lost or erroneous data packet, and set j=j+1, and return to step b1, where j is an integer greater than zero;

[0045] The expression of the pre-built priority calculation model is as follows:

[0046] ;

[0047] Where: is the priority coefficient, is a weight factor greater than zero, is an exponential function, is the number of retransmissions;

[0048] b4: Repeat the above steps b1 to b3 until j=U, then end the loop and obtain the priority coefficients of U lost or erroneous data packets.

[0049] Further, retransmitting the lost or erroneous data packets according to the priority coefficient includes:

[0050] Extracting priority coefficients of U lost or erroneous data packets;

[0051] Sort the priority coefficients of the U lost or erroneous data packets by value from large to small, and obtain the retransmission order of the U lost or erroneous data packets;

[0052] Retransmit lost or erroneous data packets according to the retransmission order.

[0053] A data transmission management system for a non-invasive ventilator, comprising:

[0054] A traffic prediction module, used to obtain current traffic characteristic data from time QM to Q, and input the current traffic characteristic data into a preset first machine learning model for predicting future traffic characteristic data, so as to obtain future traffic characteristic data in a future time period [Q+M, Q+2M], where Q and M are integers greater than zero;

[0055] An MTU optimization module, used for obtaining an initial MTU size, inputting the future traffic feature data and the initial MTU size into a preset second machine learning model for predicting an optimal MTU size, so as to obtain an optimal MTU size;

[0056] A transmission management module, used for adjusting the initial MTU size to the optimal MTU size when in the future time period [Q+M, Q+2M], and transmitting data message fragments of the data generated by the non-invasive ventilator according to the optimal MTU size;

[0057] The retransmission processing module is used to screen out the U lost or erroneous data packets and data packet characteristic data generated in the future time period [Q+M, Q+2M], obtain the priority coefficients of the U lost or erroneous data packets according to the data packet characteristic data, and retransmit the lost or erroneous data packets according to the priority coefficients, where U is an integer greater than zero.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] The present application discloses a data transmission management method and system for a non-invasive ventilator, comprising: inputting current flow characteristic data into a preset first machine learning model to obtain future flow characteristic data in a future time period; inputting the future flow characteristic data and the initial MTU size into a preset second machine learning model to obtain the optimal MTU size; when in the future time period, transmitting data message fragments of the data generated by the non-invasive ventilator according to the optimal MTU size; screening out U lost or erroneous data packets and data packet characteristic data generated, obtaining priority coefficients of U lost or erroneous data packets according to the data packet characteristic data, and retransmitting the lost or erroneous data packets according to the priority coefficient; based on the above characteristics, the present invention can dynamically adjust the MTU size, thereby improving the throughput of non-invasive ventilator data in a good network state and the transmission efficiency of data packets, and reducing the packet loss rate or retransmission amount in a network fluctuation state, which is beneficial to improving network transmission stability; in addition, after dynamically adjusting the MTU size, priority coefficients are calculated for different types of lost or erroneous data packets, and the present invention is beneficial to ensuring the timeliness of transmission of respiratory parameter data in non-invasive ventilator data. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A flow chart of a data transmission management method for a non-invasive ventilator provided by the present invention;

[0061] Figure 2 A module structure diagram of a data transmission management system for a non-invasive ventilator provided by the present invention. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] Example 1

[0064] See also Figure 1 As shown, this embodiment discloses a data transmission management method for a non-invasive ventilator, the method comprising:

[0065] S101: obtaining current traffic characteristic data from time QM to time Q, and inputting the current traffic characteristic data into a preset first machine learning model for predicting future traffic characteristic data, so as to obtain future traffic characteristic data in a future period [Q+M, Q+2M], where Q and M are integers greater than zero;

[0066] Specifically, the current traffic characteristic data includes the initial bandwidth utilization, initial delay, initial packet loss rate, initial jitter, initial data transmission frequency and initial network congestion coefficient at each moment; wherein, the initial bandwidth utilization refers to the ratio of the actual usage of the network bandwidth at the current moment to the total bandwidth, which is obtained by measuring the ratio of the actual usage of the current bandwidth to the total network bandwidth through a network monitoring tool or a traffic analyzer; the initial delay refers to the transmission time of the data (generated by the non-invasive ventilator) from the source to the destination at the current moment, which is obtained by measuring the time difference from the sending to the receiving of the data packet; the initial packet loss rate is the proportion of data packets lost in the network at the current moment, which is measured by sending a certain number of data packets and calculating the number of successful arrivals; the initial jitter refers to the variation of the network delay at the current moment, that is, the fluctuation of the delay, which is obtained by measuring the time series data of the delay and calculating its standard deviation or variance; the initial data transmission frequency refers to the frequency of sending data packets per unit time at the current moment, which is obtained by calculating the number of data packets sent per second;

[0067] The generation logic of the initial network congestion coefficient is as follows:

[0068] Acquire network congestion characterization data, wherein the network congestion characterization data includes network traffic, network bandwidth, average data packet arrival rate, and data packet service rate from time QM to Q;

[0069] The network congestion characterization data is input into the pre-built congestion coefficient calculation model to obtain the initial network congestion coefficient.

[0070] The expression of the pre-built congestion coefficient calculation model is as follows:

[0071] ;

[0072] Where: is the initial network congestion coefficient; is the network traffic at the i-th moment, that is, the amount of data passing through the network at that time point; is the network bandwidth; is the average data packet arrival rate at the i-th moment, reflecting the average number of data packets arriving at the network per unit time; is the packet service rate at the i-th moment, which indicates the average number of packets that can be served (processed) by the network per second; is the length of the time interval, that is, the duration from QM to Q;

[0073] It should be noted that the average packet arrival rate is obtained by collecting the packet arrival data at each time point or time interval within a period of time (this is usually done through network monitoring tools or by analyzing the traffic logs of the network interface) and dividing it by the observation time. The specific formula is: The packet service rate is obtained by substituting the known average processing time of each packet (seconds / packet) into the calculation formula. The calculation formula is: .

[0074] In implementation, the generation logic of the preset first machine learning model is as follows:

[0075] Acquire historical traffic feature training data, and divide the historical traffic feature training data into a traffic feature training set and a traffic feature test set, wherein the historical traffic feature training data includes current traffic feature data from time QM to time Q and future traffic feature data in the corresponding future time period [Q+M, Q+2M];

[0076] The future traffic characteristic data includes the future bandwidth utilization, future delay, future packet loss rate, future jitter, future data transmission frequency and future network congestion coefficient at each moment in the future time period [Q+M, Q+2M];

[0077] It should be noted that the current traffic feature data and future traffic feature data in the historical traffic feature training data are obtained by technicians after being collected or calculated through various network monitoring tools and calculation models (see above); the network monitoring tools include but are not limited to network traffic probes, MQopMg, PRQG MeQworkMoMiQor, etc.;

[0078] Construct a first regression network, use the current traffic feature data in the traffic feature training set as the first regression network input data, and use the future traffic feature data in the future time period [Q+M, Q+2M] in the traffic feature training set as the output data of the first regression network, train the first regression network, and obtain a trained first machine learning network;

[0079] The trained first machine learning network is model verified using the traffic feature test set, and a first machine learning network with a value less than or equal to a first preset test error threshold is output as the preset first machine learning model.

[0080] It should be noted that: the first regression network is a specific one of regression algorithm models such as polynomial regression, random forest, decision tree, support vector machine, linear regression or neural network;

[0081] It should also be noted that the MTU (Maximum Transmission Unit) size refers to the maximum allowable size of a data packet when it is transmitted in a network link (the maximum allowable size is in bytes, such as the MTU size is 1500 bytes), which includes data and protocol headers, and the current MTU size is determined based on the initial configuration or settings of the network interface; accordingly, the current MTU size is the maximum allowable size of a data packet when it is transmitted in a network link at the current moment.

[0082] S102: Obtain an initial MTU size, and input the future traffic feature data and the initial MTU size into a preset second machine learning model for predicting an optimal MTU size to obtain the optimal MTU size;

[0083] It should be understood that: the initial MTU size, in the initial stage, refers to the default MTU value set at the beginning of data transmission, which is usually pre-set by the network administrator or equipment manufacturer; when the MTU value is adjusted over time, the initial MTU size is the optimal MTU size of the previous period, obtained through actual records;

[0084] In implementation, the generation logic of the preset second machine learning model is as follows:

[0085] Acquire historical MTU size training data, and divide the historical MTU size training data into an MTU size training set and an MTU size test set, wherein the historical MTU size training data includes future traffic feature data and its corresponding optimal MTU size;

[0086] The optimal MTU size is obtained by optimizing using a preconfigured genetic algorithm;

[0087] Specifically, the optimization logic of the preconfigured genetic algorithm is as follows:

[0088] a1: Initialize the population: randomly generate an original population, which contains G individuals, each of which represents an MTU size, and G is an integer greater than zero;

[0089] a2: Fitness evaluation: Under each individual (i.e., each MTU size in the original population), detect the data retransmission amount, packet loss rate and actual network congestion coefficient of the transmission network; input the data retransmission amount and packet loss rate into the pre-built fitness function to calculate the fitness of each individual;

[0090] It should be noted that the retransmission volume (ReQraMsmiQQed Volume) refers to the amount of data that is retransmitted in the network due to data loss, damage or other reasons. The calculation formula is: , where: is the retransmission amount, is the number of retransmitted packets, is the size of the retransmitted data packet; the packet loss rate is calculated as , where: is the packet loss rate, is the number of packets lost, is the total number of data packets sent; and the actual network congestion coefficient is the same as the calculation logic of the above initial network congestion coefficient or the future initial network congestion coefficient. For details, please refer to the above related content and will not be repeated here;

[0091] The calculation formula of the pre-constructed fitness function is: ; Where: For fitness, is the amount of data retransmission, is the packet loss rate;

[0092] a3: Selection: Use the roulette method to select two individuals with high fitness in the original population as the father and mother;

[0093] The roulette wheel method is a commonly used selection method, which is used in genetic algorithms to select individuals with higher fitness to enter the next generation. It simulates the process of roulette, and each individual obtains a corresponding "roulette wheel" area according to its fitness. The higher the fitness of the individual, the larger the corresponding area, and the higher the probability of being selected.

[0094] a4: Crossover: Perform crossover operation on the father and mother to produce new individuals;

[0095] It should be noted that the crossover operation on the paternal parent and the maternal parent is implemented based on a crossover operation, and the crossover operation includes but is not limited to one of a single-point crossover, a uniform crossover or a sequential crossover, etc.;

[0096] a5: Mutation: Perform mutation operation on the new individuals to obtain S new individuals, combine the S new individuals into a new population, replace the original population with the new population, and return to step a2;

[0097] In genetic algorithms, mutation operations are used to introduce genetic diversity and prevent the algorithm from falling into a local optimum. The mutation operation on new individuals is implemented by uniform mutation or Gaussian mutation.

[0098] a6: Repeat steps a2 to a5 above until the fitness of the individuals in the original population or the new population is greater than or equal to the preset fitness threshold, or the number of iterations is greater than or equal to the preset maximum number of iterations threshold, and output the MTU size represented by the corresponding individual as the optimal MTU size;

[0099] For example: Assume that the maximum number of iterations is 100 times, and record the individual with the highest fitness in the current population and its fitness value after each iteration; if it is found that the fitness value has not changed significantly in a certain generation, it is considered that the convergence condition is met, the iteration is stopped, and the MTU size represented by the corresponding individual is output as the optimal MTU size.

[0100] Constructing a second regression network, using the future traffic feature data in the MTU size training set as input data of the second regression network, and using the optimal MTU size in the MTU size training set as output data of the second regression network, training the second regression network, and obtaining a trained second machine learning network;

[0101] The trained second machine learning network is model verified using the MTU size test set, and a second machine learning network with a value less than or equal to a preset second test error threshold is output as the preset second machine learning model.

[0102] It should be noted that: like the above-mentioned first regression network, the second regression network is a specific one of regression algorithm models such as polynomial regression, random forest, decision tree, support vector machine, linear regression or neural network.

[0103] S103: When in the future time period [Q+M, Q+2M], the initial MTU size is adjusted to the optimal MTU size, and data message fragments of the data generated by the non-invasive ventilator are transmitted according to the optimal MTU size;

[0104] It should be understood that: a data packet includes a header, a payload part and a tail, wherein the header contains control information, which is used to record information such as the source IP address and the destination IP address of the data packet to guide the transmission direction of the data packet; when the MTU size is increased, each data packet can carry more data, so the number of data packets required to transmit the same amount of data is reduced, and fewer data packets mean less header information, which can improve data transmission efficiency, but when encountering a fluctuating network, it will cause a larger packet loss rate or retransmission volume; and when the MTU size is reduced, it can help reduce the packet loss rate and retransmission volume, because smaller data packets are easier to pass through congested network links, but it also leads to an increase in the number of data packets required to transmit the same amount of data, and more data packets mean more header information, which increases network transmission pressure and reduces overall transmission efficiency;

[0105] It can be further understood that: when the network tends to be stable, the bandwidth is sufficient and the congestion is low, by appropriately increasing the MTU size, the number of data packets can be reduced, and the data throughput and data packet transmission efficiency can be improved; and when the network tends to be fluctuating, the bandwidth is insufficient and the congestion is serious, by appropriately reducing the MTU size, the packet loss rate or retransmission amount can be reduced, and the network transmission stability can be improved, which is beneficial to ensure that the non-invasive ventilator data can be transmitted to the target device in a timely and complete manner.

[0106] S104: When the future time period is [Q+M, Q+2M], U lost or erroneous data packets and data packet characteristic data are screened out, priority coefficients of the U lost or erroneous data packets are obtained according to the data packet characteristic data, and the lost or erroneous data packets are retransmitted according to the priority coefficients, where U is an integer greater than zero;

[0107] It should be noted that: the U lost or erroneous data packets are screened by using a data packet detection mechanism, and the data packet detection mechanism includes but is not limited to a checksum mechanism, a cyclic redundancy check mechanism, and a confirmation response mechanism, etc. The lost or erroneous data packets generated in the future time period [Q+M, Q+2M] can be effectively obtained by detecting the data packet detection mechanism;

[0108] In implementation, obtaining priority coefficients of U lost or erroneous data packets includes:

[0109] b1: extracting data packet characteristic data of the jth lost or erroneous data packet, including the data type and number of retransmissions of the data packet characteristic data;

[0110] Wherein, the data type (of non-invasive ventilator data) includes respiratory parameter data and device status data, the data type is specifically determined according to the identification field of the lost or erroneous data packet, and the number of retransmissions is obtained according to the actual monitoring record;

[0111] b2: Input the data type into a predefined importance function to obtain the importance value of the lost or erroneous packet. ;

[0112] The expression of the importance function is as follows:

[0113] ;

[0114] Where: Indicates data type; Indicates that the data type is non-respiratory parameter data;

[0115] b3: Input the importance value and the number of retransmissions into the pre-built priority calculation model to obtain the priority coefficient of the jth lost or erroneous data packet, and set j=j+1, and return to step b1, where j is an integer greater than zero;

[0116] The expression of the pre-built priority calculation model is as follows:

[0117] ;

[0118] Where: is the priority coefficient, is a weight factor greater than zero, is an exponential function, is the number of retransmissions;

[0119] b4: Repeat the above steps b1 to b3 until j=U, then end the loop and obtain the priority coefficients of U lost or erroneous data packets.

[0120] In implementation, retransmitting lost or erroneous data packets according to the priority coefficient includes:

[0121] Extracting priority coefficients of U lost or erroneous data packets;

[0122] Sort the priority coefficients of the U lost or erroneous data packets by value from large to small, and obtain the retransmission order of the U lost or erroneous data packets;

[0123] Retransmit lost or erroneous data packets according to the retransmission order.

[0124] By predicting the future traffic characteristic data in the future time period [Q+M, Q+2M] according to the current traffic characteristic data, and determining the optimal MTU size based on the future traffic characteristic data, the present invention can dynamically adjust the data packet size when transmitting non-invasive ventilator data, so that when the network status is in a good state, it is beneficial to improve the data throughput and the transmission efficiency of the data packet, and thus it is beneficial to improve the transmission speed of the non-invasive ventilator data; and when the network status is in a poor state, it is beneficial to improve the data throughput and the transmission efficiency of the data packet, and thus it is beneficial to reduce the packet loss rate or retransmission amount of the data packet, improve the network transmission stability, and thus it is beneficial to ensure that the non-invasive ventilator data can be transmitted to the target device in a timely and complete manner;

[0125] In addition, even after network transmission through a dynamic MTU size, there may still be a very small portion of lost or erroneous data packets. By calculating priority coefficients for different types of lost or erroneous data packets and retransmitting each lost or erroneous data packet in a retransmission order according to the priority coefficient, the present invention is conducive to ensuring the timeliness of the transmission of respiratory parameter data.

[0126] Example 2

[0127] See also Figure 2 As shown, based on the same inventive concept, this embodiment discloses a data transmission management system for a non-invasive ventilator. For details not provided in this embodiment, please refer to the relevant parts of the above embodiment 1. The system includes:

[0128] The traffic prediction module 210 is used to obtain current traffic characteristic data from time QM to Q, and input the current traffic characteristic data into a preset first machine learning model for predicting future traffic characteristic data to obtain future traffic characteristic data in a future period [Q+M, Q+2M], where Q and M are integers greater than zero;

[0129] An MTU optimization module 220 is used to obtain an initial MTU size, and input the future traffic feature data and the initial MTU size into a preset second machine learning model for predicting an optimal MTU size to obtain an optimal MTU size;

[0130] The transmission management module 230 is used to adjust the initial MTU size to the optimal MTU size when in the future time period [Q+M, Q+2M], and transmit the data message fragments of the data generated by the non-invasive ventilator according to the optimal MTU size;

[0131] The retransmission processing module 240 is used to filter out the U lost or erroneous data packets and data packet characteristic data generated in the future time period [Q+M, Q+2M], obtain the priority coefficients of the U lost or erroneous data packets according to the data packet characteristic data, and retransmit the lost or erroneous data packets according to the priority coefficients, where U is an integer greater than zero.

[0132] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters, weights and thresholds in the formula are set by technicians in this field according to actual conditions.

[0133] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired network or a wireless network. The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD) or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0134] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0135] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0136] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0137] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0138] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0139] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0140] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A data transmission management method for a non-invasive ventilator, characterized in that: The method comprises: Obtain current traffic characteristic data from time QM to time Q, and input the current traffic characteristic data into a preset first machine learning model for predicting future traffic characteristic data to obtain future traffic characteristic data in a future time period [Q+M, Q+2M], where Q and M are integers greater than zero; The current traffic characteristic data includes the initial bandwidth utilization, initial delay, initial packet loss rate, initial jitter, initial data transmission frequency and initial network congestion coefficient at each moment; The generation logic of the initial network congestion coefficient is as follows: Acquire network congestion characterization data, wherein the network congestion characterization data includes network traffic, network bandwidth, average data packet arrival rate, and data packet service rate from time QM to Q; Inputting the network congestion characterization data into a pre-built congestion coefficient calculation model to obtain an initial network congestion coefficient; The expression of the pre-built congestion coefficient calculation model is as follows: ； Where: is the initial network congestion coefficient; is the network traffic at the i-th moment; is the network bandwidth; is the average packet arrival rate at the i-th moment; is the data packet service rate at the i-th moment; is the length of the time interval; The generation logic of the preset first machine learning model is as follows: Acquire historical traffic feature training data, and divide the historical traffic feature training data into a traffic feature training set and a traffic feature test set, wherein the historical traffic feature training data includes current traffic feature data from time QM to time Q and future traffic feature data in the corresponding future time period [Q+M, Q+2M]; The future traffic characteristic data includes the future bandwidth utilization, future delay, future packet loss rate, future jitter, future data transmission frequency and future network congestion coefficient at each moment in the future time period [Q+M, Q+2M]; Construct a first regression network, use the current traffic feature data in the traffic feature training set as the first regression network input data, and use the future traffic feature data in the future time period [Q+M, Q+2M] in the traffic feature training set as the output data of the first regression network, train the first regression network, and obtain a trained first machine learning network; Using the traffic feature test set to perform model verification on the trained first machine learning network, and outputting a first machine learning network with a value less than or equal to a first preset test error threshold as a preset first machine learning model; Obtaining an initial MTU size, and inputting the future traffic feature data and the initial MTU size into a preset second machine learning model for predicting an optimal MTU size to obtain the optimal MTU size; The generation logic of the preset second machine learning model is as follows: Acquire historical MTU size training data, and divide the historical MTU size training data into an MTU size training set and an MTU size test set, wherein the historical MTU size training data includes future traffic feature data and its corresponding optimal MTU size; Constructing a second regression network, using the future traffic feature data in the MTU size training set as input data of the second regression network, and using the optimal MTU size in the MTU size training set as output data of the second regression network, training the second regression network, and obtaining a trained second machine learning network; Use the MTU size test set to perform model verification on the trained second machine learning network, and output a second machine learning network that is less than or equal to a preset second test error threshold as a preset second machine learning model; When in the future time period [Q+M, Q+2M], the initial MTU size is adjusted to the optimal MTU size, and the data message fragments of the data generated by the non-invasive ventilator are transmitted according to the optimal MTU size; When in the future time period [Q+M, Q+2M], the U lost or erroneous data packets and data packet characteristic data generated are screened out, priority coefficients of the U lost or erroneous data packets are obtained according to the data packet characteristic data, and the lost or erroneous data packets are retransmitted according to the priority coefficients, where U is an integer greater than zero.

2. The data transmission management method for a non-invasive ventilator according to claim 1, characterized in that: The optimal MTU size is obtained by optimizing using a preconfigured genetic algorithm; The optimization logic of the preconfigured genetic algorithm is as follows: a1: Initialize the population: randomly generate an original population, which contains G individuals, each of which represents an MTU size, and G is an integer greater than zero; a2: Fitness evaluation: Under each individual, detect the data retransmission amount, packet loss rate and actual network congestion coefficient of the transmission network; input the data retransmission amount and packet loss rate into the pre-built fitness function to calculate the fitness of each individual; a3: Selection: Use the roulette method to select two individuals with high fitness in the original population as the father and mother; a4: Crossover: Perform crossover operation on the father and mother to produce new individuals; a5: Mutation: Perform mutation operation on the new individuals to obtain S new individuals, combine the S new individuals into a new population, replace the original population with the new population, and return to step a2; a6: Repeat steps a2 to a5 above until the fitness of the individuals in the original population or the new population is greater than or equal to the preset fitness threshold, or the number of iterations is greater than or equal to the preset maximum number of iterations threshold, and output the MTU size represented by the corresponding individual as the optimal MTU size.

3. The data transmission management method for a non-invasive ventilator according to claim 2, characterized in that: The calculation formula of the pre-constructed fitness function is: ; Where: For fitness, is the amount of data retransmission, is the packet loss rate.

4. The data transmission management method for a non-invasive ventilator according to claim 3, characterized in that: The obtaining priority coefficients of U lost or erroneous data packets includes: b1: extracting data packet characteristic data of the jth lost or erroneous data packet, including the data type and number of retransmissions of the data packet characteristic data; b2: Input the data type into a predefined importance function to obtain the importance value of the lost or erroneous packet. ; The expression of the importance function is as follows: ; Where: Indicates data type; Indicates that the data type is non-respiratory parameter data; b3: Input the importance value and the number of retransmissions into the pre-built priority calculation model to obtain the priority coefficient of the jth lost or erroneous data packet, and set j=j+1, and return to step b1, where j is an integer greater than zero; The expression of the pre-built priority calculation model is as follows: ； Where: is the priority coefficient, is a weight factor greater than zero, is an exponential function, is the number of retransmissions; b4: Repeat the above steps b1 to b3 until j=U, then end the loop and obtain the priority coefficients of U lost or erroneous data packets.

5. The data transmission management method for a non-invasive ventilator according to claim 4, characterized in that: The retransmitting of lost or erroneous data packets according to the priority coefficient comprises: Extracting priority coefficients of U lost or erroneous data packets; Sort the priority coefficients of the U lost or erroneous data packets by value from large to small, and obtain the retransmission order of the U lost or erroneous data packets; Retransmit lost or erroneous data packets according to the retransmission order.

6. A data transmission management system for a non-invasive ventilator, characterized in that: The method is implemented based on the data transmission management method for a non-invasive ventilator according to any one of claims 1 to 5, and the system comprises: A traffic prediction module, used to obtain current traffic characteristic data from time QM to Q, and input the current traffic characteristic data into a preset first machine learning model for predicting future traffic characteristic data, so as to obtain future traffic characteristic data in a future time period [Q+M, Q+2M], where Q and M are integers greater than zero; An MTU optimization module, used for obtaining an initial MTU size, inputting the future traffic feature data and the initial MTU size into a preset second machine learning model for predicting an optimal MTU size, so as to obtain an optimal MTU size; A transmission management module, used for adjusting the initial MTU size to the optimal MTU size when in the future time period [Q+M, Q+2M], and transmitting data message fragments of the data generated by the non-invasive ventilator according to the optimal MTU size; The retransmission processing module is used to screen out the U lost or erroneous data packets and data packet characteristic data generated in the future time period [Q+M, Q+2M], obtain the priority coefficients of the U lost or erroneous data packets according to the data packet characteristic data, and retransmit the lost or erroneous data packets according to the priority coefficients, where U is an integer greater than zero.

Images