A reliability assessment method for an infusion termination device

Abstract

The present invention discloses a reliability assessment method for an end-of-infusion device. Existing reliability assessment methods for end-of-infusion devices are insufficient to timely, accurately, and comprehensively assess the reliability of end-of-infusion devices. The present invention uses test data, the fireworks algorithm, and the least squares method to estimate the parameters of the Irene model, establishing an Irene model for end-of-infusion devices. The life distribution function of the end-of-infusion device is then fitted based on the failure time data of all end-of-infusion devices after accelerated life testing. The Irene model and the life distribution function together form a reliability assessment model for the end-of-infusion device. The present invention effectively solves the problems of reliability assessment and safe operation of end-of-infusion devices within their lifespan.

Description

Technical Field

[0001] The present invention belongs to the technical field of reliability assessment of infusion termination, relates to an infusion termination device, and particularly relates to a reliability assessment method for an infusion termination device. Background Art

[0002] The end-of-infusion device is an important device among medical devices, widely used in hospitals, clinics and other places. It can timely alarm when the infusion is finished, preventing blood from flowing back and forming blood clots, which may cause medical accidents. The end-of-infusion devices on the market mainly use two methods: photoelectric counting and sensor weighing. The photoelectric principle of the end-of-infusion device has high requirements for electronic components, and often requires the end-of-infusion device to have a highly stable working condition. The stable operation of the end-of-infusion device is directly related to the safety of the infusion. This type of end-of-infusion alarm device works in an indoor environment and is constantly moving with the infusion device. The electronic components in the end-of-infusion device may fail under temperature, humidity and vibration conditions. Its main failure mode is sensor failure, and the parameter that characterizes the performance is sensor sensitivity. The end-of-infusion alarm device is used for non-manual automatic monitoring of the end of infusion. If the end-of-infusion device fails, it will cause serious medical accidents. Therefore, an accurate reliability assessment method for the end-of-infusion device is needed. However, existing end-of-infusion device reliability assessment methods typically rely on analyzing field feedback data. This method, due to the diverse use cases, results in widely varying failure data, making it impossible to reach a unified conclusion. Furthermore, field feedback data is often incomplete, unspecific, time-consuming, and expensive. Consequently, traditional end-of-infusion device assessment methods are insufficient for timely, accurate, and comprehensive reliability assessments of end-of-infusion devices. Summary of the Invention

[0003] To solve the above problems, the present invention aims to provide a reliability assessment method for an infusion termination device.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] The present invention provides a reliability assessment method for an end-of-infusion device, comprising the following steps: determining accelerated stress during an accelerated life test of the end-of-infusion device, conducting an accelerated life test of the end-of-infusion device, and assessing the reliability of the end-of-infusion device. The accelerated stresses during the accelerated life test of the end-of-infusion device are determined to be temperature, humidity, and vibration. The accelerated life test of the end-of-infusion device is performed as follows: an environmental testing chamber is placed on a triaxial vibration table, and the end-of-infusion device is placed within the environmental testing chamber. The environmental testing chamber controls temperature and humidity, and the triaxial vibration table controls vibration intensity. The end-of-infusion device undergoes an accelerated life test. During the accelerated life test, the detection sensitivity of the end-of-infusion device is measured at preset intervals until the detection sensitivity of the end-of-infusion device fails to meet preset requirements. The failure time is recorded, and the accelerated life test terminates. The end-of-infusion devices are arranged in multiple groups for the accelerated life test, each group comprising at least ten end-of-infusion devices. The end-of-infusion devices in different groups undergo accelerated life tests under different combinations of temperature, humidity, and vibration intensity, while the end-of-infusion devices in the same group undergo accelerated life tests under the same combination of temperature, humidity, and vibration intensity. The reliability assessment of end-of-infusion devices is conducted using the Irene model. Specifically, the parameters of the Irene model are estimated using the fireworks algorithm and the least squares method based on the temperature, humidity, vibration intensity, and failure time data from accelerated life tests of all end-of-infusion devices. This establishes the Irene model for end-of-infusion devices. Then, the life distribution function of the end-of-infusion device is fitted based on the failure time data from accelerated life tests of all end-of-infusion devices. The Irene model and the life distribution function together constitute the reliability assessment model for the end-of-infusion device. The fireworks algorithm is used to automatically search for initial values ​​for the least squares method. After the fireworks algorithm finds the initial values ​​for the least squares estimate, the parameters of the Irene model are estimated using the least squares method. Using the reliability assessment model for the end-of-infusion device, the temperature, humidity, and vibration intensity of the end-of-infusion device under normal operating conditions are substituted into the Irene model to obtain the characteristic life parameters of the life distribution function, thereby determining the operational reliability of the end-of-infusion device.

[0006] Preferably, when all groups of end-of-infusion devices are subjected to accelerated life tests, three levels of temperature, humidity and vibration intensity are adopted, and the three levels are respectively the minimum value, the maximum value and the average value of the minimum and maximum values.

[0007] Preferably, the Irene model expression is: Where T is the absolute temperature, RH is the relative humidity, S is the vibration intensity, k is the Boltzmann constant, ΔE is the activation energy parameter, and the parameters A, B, C and ΔE are all unknown parameters.

[0008] Preferably, in the fireworks algorithm, a preset number of random numbers are generated within the feasible domain Ω based on the temperature, humidity, vibration intensity and failure time data of all the infusion end devices undergoing accelerated life tests. Each random number xi ={A i , B i , C i , ΔE i} is regarded as a firework, and each firework represents a feasible solution in the feasible region; the objective function is set as Minf(x i ),in, Minf(x i )∈R,A i is the parameter A value of the i-th random number, B i is the parameter B value of the i-th random number, C i is the parameter C value of the i-th random number, ΔE i is the parameter ΔE value of the i-th random number; the fitness value of each firework is calculated according to the objective function. Fireworks with better fitness values ​​are given more search resources, producing more fireworks in a smaller range, making the firework position have stronger local search capabilities; the explosion radius of each firework and the number of sparks produced by the explosion are calculated according to the fitness value of the firework.

[0009] More preferably, the explosion radius of the fireworks is The number of sparks produced by the explosion is obtained by Get, where y min is the minimum value of all fireworks fitness values, y max is the maximum fitness value of all fireworks, N is the number of fireworks, is a constant used to adjust the radius of the fireworks explosion, M is also a constant used to adjust the number of sparks produced by the explosion, and ε is the minimum value of the machine. and M both range from 1 to 50, and ε = 0.1.

[0010] More preferably, in order to limit the number of fireworks that explode at the location with good fitness value and the number of fireworks that do not produce excessive sparks at the location with poor fitness value, the following formula is used to calculate S i Make corrections and get:

[0011]

[0012] Wherein, round is a rounding function based on the rounding principle, a and b are both constants, and a<b<1.

[0013] More preferably, in order to increase the comprehensiveness of the solution to the Irene model, the fireworks algorithm introduces a mutation operator e to generate Gaussian mutation sparks of the Irene model, and at the same time, the Irene model is position mapped. The Gaussian mutation spark is generated by the following method: First, a firework x is randomly selected from the firework population. i , and then randomly select a certain number of dimensions for the fireworks to perform Gaussian mutation; among them, for x iThe operation of Gaussian mutation on a certain dimension k is: x ik is x i The value of dimension k, e~N(1,1), N(1,1) represents a Gaussian distribution with a mean of 1 and a variance of 1. The position mapping rule of the Irene model is as follows: where x UB，k and x LB，k are the upper and lower boundaries of the solution space of the Irene model in dimension k, respectively, and % represents a modular operation. In order to pass the excellent information in the fireworks population of the Irene model to the next generation population, after the explosion sparks and Gaussian mutation sparks are generated, a preset number of individuals are selected from the candidate set as the fireworks of the next generation, as follows: Assuming that the candidate set of the Irene model is K and the fireworks population size is N, the individual with the smallest fireworks fitness value in the candidate set is deterministically selected as the next generation of fireworks, and the roulette wheel method is used to select the remaining N-1 fireworks from the candidate set. The roulette wheel method is as follows: for candidate x i , the probability of being selected Among them, R(x i )=∑ xj∈K d(x i -x j ), x j Divide by x i For candidates other than i -x j ) is x i to x j distance.

[0014] The present invention has the beneficial effects:

[0015] This invention uses test data, the fireworks algorithm, and the least squares method to estimate the parameters of the Irene model, establishing the Irene model for end-of-infusion devices. The life distribution function of the end-of-infusion devices is then fitted based on the failure time data from accelerated life tests of all end-of-infusion devices. The Irene model and the life distribution function together form a reliability assessment model for the end-of-infusion device. The reliability model established by this invention can accurately assess the operational reliability of end-of-infusion devices, effectively addressing the lack of stability in end-of-infusion devices and providing clear reliability performance indicators for operating end-of-infusion devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0017] The present invention will be further described below with reference to the accompanying drawings.

[0018] like Figure 1 As shown, a reliability assessment method for an end-of-infusion device comprises the following steps: determining accelerated stress during an accelerated life test of the end-of-infusion device, conducting an accelerated life test of the end-of-infusion device, and assessing the reliability of the end-of-infusion device. During the daily operation of the end-of-infusion device, the electronic components within the end-of-infusion device are exposed to air, affected by ambient temperature, and exposed to moisture in the environment for a long period of time. Furthermore, the end-of-infusion device is constantly moved with the infusion site, subjecting it to long-term vibration stress. Therefore, during the reliability assessment process of the present invention, the accelerated stresses of the accelerated life test are determined to be temperature, humidity, and vibration. To expose the shortcomings of the end-of-infusion device as quickly and distinguishably as possible, the accelerated life test was designed. Temperature and humidity were simulated using the temperature monitoring module and humidity control module within the environmental test chamber, respectively. Vibration stress was simulated using a triaxial vibration table. Specifically, the environmental test chamber was placed on the triaxial vibration table, and the end-of-infusion device was placed within the environmental test chamber. The environmental test chamber controlled the temperature and humidity, and the triaxial vibration table controlled the vibration intensity. The accelerated life test was conducted on the end-of-infusion device. During the accelerated life test, the detection sensitivity of the end-of-infusion device was measured every 48 hours until the detection sensitivity of the end-of-infusion device failed to meet the preset requirements. The failure time was recorded, and the accelerated life test was terminated. The end-of-infusion devices were set up for accelerated life testing in multiple groups, each group containing at least ten end-of-infusion devices. End-of-infusion devices in different groups were tested under different combinations of temperature, humidity, and vibration intensity, and end-of-infusion devices in the same group were tested under the same combination of temperature, humidity, and vibration intensity. The reliability assessment of the end-of-infusion device is carried out using the Irene model. Specifically, based on the temperature, humidity, vibration intensity and failure time data of the accelerated life test of all end-of-infusion devices, the parameters of the Irene model are estimated using the fireworks algorithm and the least squares method, and the Irene model of the end-of-infusion device is established. Then, based on the failure time data after the accelerated life test of all end-of-infusion devices, the life distribution function of the end-of-infusion device is fitted. Therefore, the Irene model and the life distribution function together constitute a reliability assessment model of the end-of-infusion device. Among them, the fireworks algorithm is used for automatic search of the initial value of the least squares method. After the fireworks algorithm searches for the initial value of the least squares estimate, the parameters of the Irene model are estimated using the least squares method. With the help of the reliability assessment model of the end-of-infusion device, the accelerated stress (temperature, humidity and vibration intensity) of the end-of-infusion device under normal working conditions is substituted into the Irene model to obtain the life characteristic parameters of the life distribution function under accelerated stress, and the working reliability of the end-of-infusion device can be obtained, so as to grasp the healthy operating status of the end-of-infusion device and replace it before it fails, thereby ensuring the reliability of the end-of-infusion device in normal working environments such as hospitals and clinics, improving the operating stability of the end-of-infusion device, and reducing the occurrence of medical accidents caused by the infusion process.

[0019] As a preferred embodiment, when all groups of end-of-infusion devices were subjected to accelerated life tests, three levels of temperature, humidity, and vibration intensity were used, namely, minimum, maximum, and the average of the minimum and maximum values. In this embodiment, the minimum temperature was 45°C (slightly higher than the normal ambient temperature) and the maximum temperature was 85°C. The test temperatures of the end-of-infusion devices were 45°C, 65°C, and 85°C, the humidity was 72% RH, 84% RH, and 96% RH, and the vibration intensity was 0.8g. 2 / Hz, 0.5g 2 / Hz and 0.2g 2 / Hz, and a total of 15 groups of infusion end devices were set up for accelerated life tests; the temperature, humidity and vibration intensity of each group of infusion end devices during the accelerated life tests are listed in Table 1.

[0020]

[0021] As a preferred embodiment, the Irene model expression is: Where T is absolute temperature, RH is relative humidity, S is vibration intensity, k is the Boltzmann constant, ΔE is the activation energy parameter, and parameters A, B, C, and ΔE are all undetermined. The Irene model is well suited for reliability assessment models involving multiple stresses, such as temperature, humidity, and vibration intensity.

[0022] As a preferred embodiment, in the fireworks algorithm, a preset number of random numbers are generated within the feasible domain Ω based on the temperature, humidity, vibration intensity and failure time data of all infusion end devices undergoing accelerated life tests. Each random number x i ={A i , B i , C i , ΔE i} is regarded as a firework, and each firework represents a feasible solution in the feasible region; the objective function is set as Minf(x i ),in, Minf(x i )∈R,A i is the parameter A value of the i-th random number, B i is the parameter B value of the i-th random number, C i is the parameter C value of the i-th random number, ΔE i is the parameter ΔE value of the i-th random number; the fitness value of each firework is calculated according to the objective function. Fireworks with better fitness values ​​are given more search resources, producing more fireworks in a smaller range, making the firework position have stronger local search capabilities; the explosion radius of each firework and the number of sparks produced by the explosion are calculated according to the fitness value of the firework.

[0023] As a more preferred embodiment, the explosion radius of the fireworks is The number of sparks produced by the explosion is obtained by Get, where y min is the minimum value of all fireworks fitness values, y max is the maximum fitness value of all fireworks, N is the number of fireworks, is a constant used to adjust the radius of the fireworks explosion, M is also a constant used to adjust the number of sparks produced by the explosion, and ε is the minimum value of the machine to avoid zero operations. and M both range from 1 to 50, and ε = 0.1.

[0024] As a more preferred embodiment, in order to limit the number of fireworks that explode at the location with good fitness value, and to prevent the location with poor fitness value from producing excessive spark particles, the following formula is used to calculate S i Make corrections and get:

[0025]

[0026] Wherein, round is a rounding function based on the rounding principle, a and b are both constants, and a<b<1.

[0027] As a more preferred embodiment, in order to increase the comprehensiveness of the solution of the Irene model, the fireworks algorithm introduces a mutation operator e to generate fireworks mutation sparks of the Irene model, namely Gaussian mutation sparks, and simultaneously performs position mapping on the Irene model. The Gaussian mutation sparks are generated by the following method: First, a firework x is randomly selected from the fireworks population. i , and then randomly select a certain number of dimensions for the firework (i.e. in A i , B i , C i , ΔE i Select several of the four dimensions) to perform Gaussian mutation; among them, for x i The operation of Gaussian mutation on a certain dimension k is: x ik is x i The value of e in dimension k follows the N(1, 1) distribution, where N(1, 1) represents a Gaussian distribution with a mean of 1 and a variance of 1. The position mapping rule of the Irene model is as follows: where x UB，k and x LB，kare the upper and lower boundaries of the solution space of the Irene model in dimension k, respectively, and % represents the modular operation. In order to pass the excellent information in the fireworks population of the Irene model to the next generation population, after the explosion sparks and Gaussian mutation sparks are generated, a preset number of individuals are selected from the candidate set (including fireworks, explosion sparks, and Gaussian mutation sparks) as the fireworks of the next generation, as follows: Assuming that the candidate set of the Irene model is K, and the size of the fireworks population is N, the individual with the smallest fireworks fitness value in the candidate set is deterministically selected to the next generation as a firework (elite strategy), and the roulette method is used to select the remaining N-1 fireworks from the candidate set. The roulette method is as follows: for candidate x i , the probability of being selected Among them, R(x i )=∑ xj∈K d(x i -x j ), x j Divide by x i For candidates other than i -x j ) is x i to x j distance.

[0028] The above are only preferred embodiments of the present invention. For those skilled in the art, the present invention may be modified and varied in various ways. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A reliability assessment method for an end-of-infusion device, comprising the following steps: determining accelerated stress during an accelerated life test of the end-of-infusion device, conducting an accelerated life test of the end-of-infusion device, and assessing the reliability of the end-of-infusion device; wherein the accelerated stress during the accelerated life test of the end-of-infusion device is determined to be temperature, humidity, and vibration; and wherein: The accelerated life test of the end-of-infusion device is as follows: an environmental test box is placed on a three-axis vibration table, and the end-of-infusion device is placed in the environmental test box. The temperature and humidity are controlled by the environmental test box, and the vibration intensity is controlled by the three-axis vibration table. The end-of-infusion device is subjected to an accelerated life test. During the accelerated life test, the detection sensitivity of the end-of-infusion device is measured at each preset time interval until the detection sensitivity of the end-of-infusion device does not meet the preset requirements, the failure time is recorded, and the accelerated life test is ended; wherein, multiple groups of end-of-infusion devices are set to perform accelerated life tests, and each group includes at least ten end-of-infusion devices. End-of-infusion devices of different groups are subjected to accelerated life tests under different combinations of temperature, humidity and vibration intensity, and end-of-infusion devices of the same group are subjected to accelerated life tests under the same combination of temperature, humidity and vibration intensity; the reliability evaluation of the end-of-infusion device is performed using the Irene model, which has The present invention is characterized in that: based on the temperature, humidity, vibration intensity and failure time data of all end-of-infusion devices subjected to accelerated life tests, the parameters of the Irene model are estimated using the fireworks algorithm and the least squares method, and the Irene model of the end-of-infusion device is established. Then, based on the failure time data after the accelerated life tests of all end-of-infusion devices, the life distribution function of the end-of-infusion device is fitted, so that the Irene model and the life distribution function together constitute a reliability assessment model of the end-of-infusion device. Among them, the fireworks algorithm is used for automatic initial value search of the least squares method. After searching the initial value of the least squares estimate, the parameters of the Irene model are estimated using the least squares method. With the help of the reliability assessment model of the end-of-infusion device, the temperature, humidity and vibration intensity of the end-of-infusion device under normal working conditions are substituted into the Irene model to obtain the life characteristic parameters of the life distribution function, and then the working reliability of the end-of-infusion device is obtained. The Irene model expression is: ,in, is the absolute temperature, RH is the relative humidity, is the vibration intensity, is the Boltzmann constant, is the activation energy parameter, parameters A, B, C and All parameters are pending.

2. The reliability assessment method for an infusion termination device according to claim 1, characterized in that: When the infusion end devices of all groups were subjected to accelerated life tests, three levels of temperature, humidity and vibration intensity were used, namely the minimum value, the maximum value and the average of the minimum and maximum values.

3. A reliability assessment method for an infusion termination device according to claim 1 or 2, characterized in that: In the fireworks algorithm, based on the temperature, humidity, vibration intensity and failure time data of the accelerated life test of all the infusion end devices, the feasible domain Generate a preset number of random numbers, each random number Considered as a firework, each firework represents a feasible solution in the feasible region; the objective function is set to ,in, , , is the parameter A value of the i-th random number, is the parameter B value of the i-th random number, is the parameter C value of the i-th random number, is the parameter of the i-th random number value; the fitness value of each firework is calculated according to the objective function. Fireworks with better fitness values ​​are given more search resources, producing more fireworks in a smaller range, making the firework position have stronger local search capabilities; the explosion radius of each firework and the number of sparks produced by the explosion are calculated according to the fitness value of the firework.

4. The reliability assessment method for an infusion termination device according to claim 3, characterized in that: The explosion radius of said firework is determined by The number of sparks produced by the explosion is obtained by Get, among them, is the minimum value of all fireworks fitness values, is the maximum value of all fireworks fitness values, is the number of fireworks, is a constant used to adjust the radius of the fireworks explosion, and M is also a constant used to adjust the number of sparks produced by the explosion. is the minimum value of the machine, Both M and α are in the range of 1-50. .

5. The reliability assessment method for an infusion termination device according to claim 3, characterized in that: In order to limit the fireworks position with good fitness value from exploding too many sparks, and at the same time, the fireworks position with poor fitness value will not produce spark particles, the following formula is used to Make corrections and get: in, is the rounding function based on the rounding principle, and are constants, and .

6. The reliability assessment method for an infusion termination device according to claim 3, characterized in that: In order to increase the comprehensiveness of the solution to the Irene model, the Fireworks algorithm introduces a mutation operator Generate Gaussian mutation sparks of the Irene model and map the position of the Irene model at the same time; Gaussian mutation sparks are generated by the following method: First, randomly select a firework from the firework population , and then randomly select a certain number of dimensions for the fireworks to perform Gaussian mutation; The operation of Gaussian mutation on a certain dimension k is: , for The value in dimension k, represents a Gaussian distribution with a mean of 1 and a variance of 1; the position mapping rule of the Irene model is as follows: ,in are the upper and lower boundaries of the solution space of the Irene model in dimension k, Represents modular operation; in order to pass the excellent information in the fireworks population of the Irene model to the next generation population, after the explosion sparks and Gaussian mutation sparks are generated, a preset number of individuals are selected from the candidate set as the fireworks of the next generation, as follows: Assuming that the candidate set of the Irene model is K, the fireworks population size is N, the individual with the smallest fireworks fitness value in the candidate set is deterministically selected to be the next generation of fireworks, and the roulette method is used to select the remaining N-1 fireworks from the candidate set; the roulette method is as follows: for the candidate , the probability of being selected ,in, To remove Other candidates, for arrive distance.

Images