Control method for sterilizer by artificial neural network and genetic algorithm

KR103024886B1Active Publication Date: 2026-09-29TEST ONE CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
KR1020230176435
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2023-12-07
Publication Date
2026-09-29
Estimated Expiration
2043-12-07

Smart Images

  • Figure 112023137243155-PAT00001_ABST
    Figure 112023137243155-PAT00001_ABST
Patent Text Reader

Abstract

The present invention provides a sterilizer control method based on an artificial neural network and a genetic algorithm, comprising: (a) a data collection step for collecting the initial state of the sterilizer; (b) a prediction model setting step for setting a preparation time to reach a usable stage of the sterilizer based on the data collected in step (a); and (c) a step for controlling the sterilizer according to the prediction model set in step (b).
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] The present invention relates to a method for controlling a sterilizer, and specifically, to a method for reducing preparation time based on an artificial neural network and a genetic algorithm. Background Technology

[0002] Generally, various medical instruments are stored and used after sterilization, and methods such as using chemicals or high-temperature steam have been proposed for this purpose. While medical instruments that are not used directly on the human body can be sterilized using chemicals, most medical instruments used directly on the human body are currently sterilized using high-temperature steam.

[0003] Conventional medical steam sterilizers using high-temperature steam are equipped with a chamber formed with a sterilization chamber to accommodate medical instruments to be sterilized, a heater placed at the bottom of the inner side of the chamber, a control unit for controlling various electrical components, and a door for opening and closing the chamber.

[0004] When room temperature water and the object to be sterilized are placed directly into the sterilization chamber of the chamber and the heater is operated to generate steam, the object to be sterilized is sterilized by high-temperature steam.

[0005] At this time, a process of creating a high-temperature, high-pressure steam state is carried out through a sterilization preparation stage (preheating and adjustment) before performing sterilization. However, there is a problem where the preparation stage time is excessive compared to the actual sterilization time, so optimization of the sterilizer's control is required. Prior art literature

[0006] Republic of Korea Published Patent No. 10-2015-0136862 Date of publication: December 08, 2015 The problem to be solved

[0007] The objective of the present invention is to provide a control method for a sterilizer that can reduce preparation time based on artificial neural networks and genetic algorithms. means of solving the problem

[0008] A sterilizer control method based on an artificial neural network and a genetic algorithm according to one embodiment for solving the above-mentioned problem is a control method for shortening the preparation time of a sterilizer, comprising: (a) a data collection step for collecting the initial state of the sterilizer; (b) a prediction model setting step for setting a preparation time to reach a usable stage of the sterilizer based on the data collected in step (a); and (c) a step of controlling the sterilizer according to the prediction model set in step (b).

[0009] At this time, the above step (a) may include: (a-1) a step of measuring the temperature and pressure of the internal chamber of the sterilizer; (a-2) a step of measuring the temperature of the steam supply of the sterilizer; and (a-3) a step of measuring the temperature of the heater of the sterilizer.

[0010] In addition, step (b) above can derive a sterilizer control algorithm by simulating a control method regarding the test design using an artificial neural network method based on the initially measured temperature and pressure of the inner chamber, the temperature of the steam supply, and the temperature of the heater.

[0011] Meanwhile, in step (b) above, the test design can perform simulations on a total of 45 data sets with 9 levels of P gain values ​​and 5 levels of I gain values ​​in the PI gain values.

[0012] In addition, each PI gain value of the pressure controller of the sterilizer is collected according to the experiment, and by applying the collected process variable, the PI gain, to a neural network algorithm, a dataset for the PI gain that is reached in the minimum time can be derived. Effects of the invention

[0013] According to the artificial neural network and genetic algorithm-based sterilizer control method of the present invention, the time to reach the sterilization preparation stage before performing sterilization is shortened, thereby reducing the overall sterilizer operating time. Specifically, a prediction model for the sterilization preparation stage time can be constructed using a small amount of data set (45).

[0014] In addition, by linking with a genetic algorithm, a PI Gain optimization algorithm can be derived, making it possible to minimize the total process time. Brief explanation of the drawing

[0015] FIG. 1 is a flowchart illustrating a sterilizer control method based on an artificial neural network and a genetic algorithm according to the present invention. Figure 2 is a block diagram of a sterile sterilizer controller. Figure 3 schematically illustrates the process of deriving an optimal control method by applying a genetic algorithm to a sterile sterilizer. Figure 4 is an example of a data set and the process results of a test design to which the data set was applied. Figure 5 illustrates an artificial neural network-based prediction model for the sterilization preparation step. Figure 6 illustrates the parameter optimization process within an artificial neural network model using grid search. Figure 7 illustrates the process of deriving optimal variables by applying a genetic algorithm. Figure 8 is a graph illustrating the time required for the sterilization preparation step and the sterilization execution step of a sterilizer. Specific details for implementing the invention

[0016] Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Note that in the accompanying drawings, identical components are indicated by the same reference numerals whenever possible. Furthermore, detailed descriptions of known functions and configurations that may obscure the essence of the invention will be omitted. For the same reason, some components in the accompanying drawings may be exaggerated, omitted, or schematically depicted.

[0017] FIG. 1 is a flowchart illustrating a sterilizer control method based on an artificial neural network and a genetic algorithm according to the present invention; FIG. 2 is a block diagram of a sterile sterilizer controller; FIG. 3 is a diagram schematically illustrating the process of deriving an optimal control method by applying a genetic algorithm to a sterile sterilizer; FIG. 4 is a diagram illustrating a data set and the process results of a test design applying the data set; FIG. 5 is a diagram illustrating a prediction model of the sterilization preparation stage based on an artificial neural network; FIG. 6 is a diagram illustrating the parameter optimization process within an artificial neural network model using grid search; FIG. 7 is a diagram illustrating the process of deriving optimal variables by applying a genetic algorithm; and FIG. 8 is a graph illustrating the time required for the sterilization preparation stage and the sterilization stage of the sterilizer.

[0018] Referring to FIGS. 1 to 8, the sterilizer control method based on an artificial neural network and a genetic algorithm according to the present invention is a control method for shortening the preparation time of a sterilizer (1000), and includes a data collection step (S100) for collecting an initial state of a sterilizer (1000), a prediction model setting step (S200) for setting a preparation time to reach a usable stage of a sterilizer (1000) based on the collected data, and a step (S300) for controlling the sterilizer according to the prediction model that has been set.

[0019] First, a sterilizer (1000) will be described with reference to FIG. 2. The sterilizer (1000), which is the subject of control in the present invention, is a device that performs sterilization by applying high temperature and high pressure to a subject to sterilization, and includes a pressure controller (100), a steam generator (200), a steam temperature controller (300), a water pump (400), a sterilization chamber (500), and a chamber temperature controller (600).

[0020] When the sterilizer (1000) is operated, the temperature and pressure corresponding to the sterilization conditions are not applied immediately; instead, the temperature and pressure corresponding to the sterilization conditions are prepared through a preheating and adjustment process. As described in FIG. 8, it can be seen that a significant amount of time is spent on preparation during the operation time of the sterilizer (1000). Typically, when the sterilizer (1000) is operated for 1 hour to perform sterilization, the actual time for sterilization is about 15 minutes, and the time required to prepare the sterilization conditions is 30 minutes.

[0021] The present invention provides a method for optimizing a pressure controller (100), a steam temperature controller (300), and a chamber temperature controller (600) by controlling them using an artificial neural network and genetic algorithm-based technique to minimize the time required for the sterilization preparation step.

[0022] The present invention derives an optimal PI Gain based on the initial value of the measured sterilization chamber (500), taking into account that the sterilization preparation time varies depending on the initial state of the sterilizer (1000).

[0023] In step S100, the initial state of the sterilizer is collected. That is, the temperature and pressure of the sterilization chamber (500), which is inside the sterilizer, are measured and received. If necessary, the initial values ​​of the pressure controller (100), steam generator (200), steam temperature controller (300), and water pump (400) may be further received.

[0024] Step S100 may be performed in a microcontroller that controls the sterilizer (1000), or in a separate computer terminal for control.

[0025] The control computer terminal can be any terminal capable of implementing artificial neural networks and genetic algorithms in software.

[0026] In step S200, based on the measured temperature and pressure of the sterilization chamber (50), the temperature of the steam supply, and the temperature of the heater, a control method for the test design is simulated using an artificial neural network to derive a sterilizer control algorithm.

[0027] Specifically, experimental data is collected according to changes in the PI Gain of the sterilizer pressure controller to perform DOE-based experimental design.

[0028] The test design performs simulations on a total of 45 data sets, with 9 levels of P gain values ​​and 5 levels of I gain values ​​in the PI gain values.

[0029] Specifically,

[0030] A total of 45 datasets are collected, including 9 levels of P Gain (0.1–4.0, 0.5 step) and 5 levels of I Gain (10–90, 20 step), and data collection is completed through 7 repeated experiments for verification.

[0031] The collected data information is as follows.

[0032] Process Variable (2): PI Gain

[0033] Initial state (4): Vessel temperature and pressure, steam temperature, plate temperature

[0034] Process Result (1): Sterilization Preparation Step Time

[0035] Each PI gain value of the pressure controller of the sterilizer is collected according to the experiment, and the collected process variable, the PI gain, is applied to a neural network algorithm to derive a dataset for the PI gain that is reached in the minimum time.

[0036] More specifically,

[0037] An artificial neural network-based sterilization preparation step time prediction model can be established using initial state data and PI Gain.

[0038] The construction of an artificial neural network-based sterilization preparation step time prediction model has the following embodiments.

[0039] Input data (6): Process variables + initial state

[0040] Output data (1): Process result (sterilization preparation step time)

[0041] If necessary, additional optimization of the artificial neural network model's internal parameters can be performed using grid search.

[0042] In other words, parameter optimization based on validation errors (MAE, MAPE) through additional collected data is possible (see Fig. 6).

[0043] Genetic algorithms can be further integrated into artificial neural networks for optimal control of the preparation phase.

[0044] As shown in Fig. 7, a technique for deriving the optimal PI Gain for minimizing the sterilization preparation step time based on a Genetic Algorithm (GA) can be applied.

[0045] - Genetic Algorithm: An optimization technique that mimics the evolutionary process of nature, which derives optimal variables by repeating the processes of parental selection, crossover, and mutation (generational repetition),

[0046] By replacing the fitness test model, which is a method for selecting parents of the next generation, with an artificial neural network, a method for deriving the optimal PI Gain that can achieve the minimum sterilization preparation step time based on the artificial neural network can be utilized.

[0047] The embodiments of the invention disclosed in this specification and drawings are provided merely as specific examples to facilitate the explanation of the technical content of the invention and to aid in understanding the invention, and are not intended to limit the scope of the invention. It is obvious to those skilled in the art that other variations based on the technical concept of the invention are possible in addition to the embodiments disclosed herein. Explanation of the symbols

[0048] 1000 : Sterilizer 100 : Pressure controller 200: Steam generator 300: Steam temperature controller 400: Water pump 500: Sterilization chamber 600: Chamber temperature controller

Claims

Claim 1 A control method for reducing the preparation time of a sterilizer comprises: (a) a data collection step for collecting the initial state of the sterilizer; (b) a prediction model setting step for setting a preparation time to reach a usable stage of the sterilizer based on the data collected in step (a); and (c) a step of controlling the sterilizer according to the prediction model set in step (b), wherein step (a) comprises: (a-1) a step of measuring the internal chamber temperature and pressure of the sterilizer; (a-2) a step of measuring the temperature of the steam supply unit of the sterilizer; A method for controlling a sterilizer based on an artificial neural network and genetic algorithm, comprising: (a-3) a step of measuring the heater temperature of the sterilizer; wherein step (b) derives a sterilizer control algorithm by simulating a control method regarding a test design using an artificial neural network method based on the initially measured temperature and pressure of the inner chamber, the temperature of the steam supply, and the temperature of the heater; wherein step (b) performs simulations on a total of 45 data sets in the test design, with the P gain value having 9 levels and the I gain value having 5 levels, and each PI gain value of the pressure controller of the sterilizer is collected according to the experiment, and the collected process variable PI gain is applied to a neural network algorithm to derive a data set for the PI gain that is reached in the minimum time. Claim 2 delete Claim 3 delete Claim 4 delete Claim 5 delete

Citation Information

Patent Citations

  • Apparatus and method for continuous infectious waste sterilization with enhanced automatic treatment control function and worker safety

    KR1020230047262A

  • Sterilization system for culture medium preparation

    CN115671328A

  • Data information processing method in working process of hydrogen peroxide sterilizer

    CN116627774A

  • Method for operation scheduling of heater and cooler based on artificial intelligence

    KR102209858B1