Operation method of fabric cleaning and disinfecting machine for controlling multiple drug-resistant bacteria

Abstract

The invention discloses an operation method of a fabric cleaning and disinfecting machine for controlling multiple drug-resistant bacteria, and particularly relates to the technical field of fabric cleaning and disinfecting. Obtaining the pollution intensity grade, the fabric material type and the target strain spectrum information of the target fabric, and constructing a pollution source information set; calling a preset disinfection response parameter model, and determining washing temperature, washing time and disinfectant type and concentration; based on the fabric type and the temperature resistance and chemical resistance of the fabric, a matched cleaning program set is retrieved from a database and dynamically adjusted; the water temperature and the detergent concentration are monitored in real time in the washing process, and washing parameters are adjusted according to the feedback control model; after cleaning is finished, the temperature deviation in the washing stage is extracted and analyzed with the critical inhibiting and killing threshold value of the target strain, and the disinfection response model is updated based on the analysis result; according to the method disclosed by the invention, the efficient killing of multiple drug-resistant bacteria and the self-adaptive optimization of the disinfection process can be realized, and the method is suitable for fabric treatment requirements in the fields of medical treatment, old-age care and public health.

Description

Technical Field

[0001] The present invention relates to the technical field of fabric cleaning and disinfection, and in particular to an operating method of a fabric cleaning and disinfecting machine for controlling multi-drug resistant bacteria. Background Art

[0002] With increasing demands for fabric hygiene in hospitals, nursing homes, and quarantine facilities, controlling multidrug-resistant bacteria (such as methicillin-resistant Staphylococcus aureus (MRSA), Acinetobacter baumannii, and Pseudomonas aeruginosa) on fabrics has become a critical component of infection prevention and control. Currently, commonly used household and industrial laundry equipment suffers from issues such as insufficient washing temperatures, imprecise disinfectant control, and inappropriate program settings. These issues make it difficult to effectively kill multidrug-resistant bacteria and even pose risks of cross-contamination and bacterial carryover.

[0003] Fabrics used in river valleys, humid environments, or semi-enclosed environments, in particular, carry higher microbial loads, placing stricter demands on the temperature, duration, and disinfection methods used during the washing process. While existing technologies have proposed methods such as high-temperature washing and disinfectant injection, these methods generally fail to incorporate differentiated controls based on fabric material properties, contamination intensity levels, and the tolerance of target bacterial species. This results in inconsistent cleaning and disinfection results, inefficient machine resource utilization, and high maintenance costs.

[0004] Therefore, it is urgent to propose a method for operating a fabric cleaning and disinfection machine that can intelligently control cleaning and disinfection parameters under the background of multiple microbial resistance characteristics, and take into account both fabric protection and energy-saving effects, so as to achieve efficient control of multidrug-resistant bacteria and safe treatment of fabrics. Summary of the Invention

[0005] The object of the present invention is to provide a method for operating a fabric cleaning and disinfecting machine for controlling multidrug-resistant bacteria, so as to overcome the shortcomings of the background technology.

[0006] In order to achieve the above-mentioned object, the present invention provides the following technical solution: a method for operating a fabric cleaning and disinfecting machine for controlling multidrug-resistant bacteria, comprising: S100: Obtain the pollution intensity level, fabric material type, and target bacterial species spectrum of the target fabric, and construct a pollution source information set P; S200: Based on the pollution source information set P, a preset disinfection response parameter model is called to determine the washing temperature T, washing time t, disinfectant type and concentration parameter C; S300: Based on the fabric type and its temperature and chemical resistance, a matching washing program set W is retrieved from the washing machine database and dynamically adjusted according to the disinfection requirements; S400: During the main washing process, the water temperature and detergent concentration are monitored in real time, and the washing parameters are adjusted according to the preset feedback control model; S500: extract and analyze the dynamic temperature deviation value of the washing stage during the washing and disinfection process and the critical inhibition and killing threshold of the target bacteria in the residual bacteria detection data, and automatically update the disinfection response parameter model of the fabric washing according to the analysis results.

[0007] Preferably, in S200, the disinfection response parameter model is a multivariate fitting model constructed based on a pollution source information set, wherein the pollution source information set includes pollution intensity level, fabric material type, and target bacterial species spectrum information, and the model is constructed by the following steps: The historical fabric cleaning sample data was collated, and the pollution level, material type, and bacterial species information corresponding to each batch were extracted as input variables; Extract the actual operating parameters of the corresponding batch, including washing temperature, action time, and disinfectant type and concentration, as the target output variables; Perform multidimensional regression modeling on input and output data to construct washing temperature prediction function, time prediction function and disinfectant concentration recommendation function; Introducing tolerance adjustment mechanisms and weight factors into the modeling process to control the acceptable floating range of temperature, time, and disinfectant concentration, as well as the weighted impact of each input variable; The model results are cross-validated and corrected through error feedback to achieve dynamic adaptation of disinfection response parameters.

[0008] Preferably, in S300, the dynamic adjustment according to the disinfection requirement includes: Extracting the target washing temperature, action time, and disinfectant type and concentration output by the disinfection response parameter model; Screening the standard programs that match the fabric material type in the preset cleaning program set to form an initial candidate program set; Adjust the temperature rise curve, disinfection section holding time, and disinfectant injection timing in the candidate program according to the disinfection requirement level of the target bacterial species; Increase the program setting temperature without exceeding the upper temperature limit of the fabric; Dynamically adjust the number and rhythm of flushing sections to meet the requirements of different types of disinfectants for residual suppression; The adjusted program parameters are combined to form the final operating program for the execution of this fabric cleaning and disinfection task.

[0009] Preferably, in S400, adjusting the washing parameters according to the preset feedback control model includes: A composite feedback control model comprising fuzzy control and proportional-integral-differential control is constructed, wherein the model uses water temperature, detergent concentration, and disinfection time acquired in real time during the washing process as input variables; Set target values ​​for washing parameters, including target water temperature, target detergent concentration, and desired disinfection time; Calculate the deviation between the current parameter and the target value, and determine the parameter deviation level through the fuzzy reasoning module; Based on the deviation level, the corresponding adjustment command direction is output and input into the proportional integral derivative controller to generate a continuous control signal; The control signal is used to adjust the heater output, detergent delivery rate, stirring frequency or washing phase time to dynamically correct deviations.

[0010] Preferably, in S500, the step of calculating the dynamic temperature deviation value includes: During the main washing process, the current water temperature data is obtained at a fixed sampling period and recorded to form a time series: ;in, to represents the time node from the start of washing to the completion of disinfection, and n represents the total number of time points; at the same time, the target constant temperature value output by the disinfection response parameter model during this operation is recorded, which is recorded as ; Based on the actual temperature at each sampling point With target temperature The difference between the two is used to calculate the instantaneous temperature deviation series : ; Then, the deviation of the whole process is normalized and averaged to obtain the dynamic temperature deviation value ΔT_avg of this washing stage, which is recorded as: .

[0011] Preferably, the steps for calculating the critical killing threshold of the target bacterial species are as follows: After the fabric is washed and dried, the residual bacterial data after this run is collected and the survival status of the target bacteria species in the residual sample after washing is identified; Set a target bacteria before running The initial bacterial load was ; After washing, the residual load of the bacteria was detected to be ; This operation is for The actual killing rate is defined as: ; Set recommended inhibition and killing thresholds for each type of common multidrug-resistant bacteria in advance Each target bacteria The actual inhibition rate is compared with the recommended threshold to form a difference , the expression is: ; For all target bacterial species The critical inhibition and killing threshold of the target bacterial species was obtained by taking a weighted average and combining the weight of the hazard level of each bacterial species.

[0012] Preferably, the dynamic temperature deviation value and the critical inhibition threshold are converted into a comprehensive feature vector, and the comprehensive feature vector is used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the error score value label of the disinfection response parameter model for the current operation control performance as the prediction target, and takes minimizing the sum of the prediction errors of the error score value labels of all disinfection response parameter models for the current operation control performance as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The error score value of the disinfection response parameter model for the current operation control performance is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0013] Preferably, when the error score E_model of the disinfection response parameter model for the control performance of this operation exceeds the set threshold E_thresh, it is determined that there is a configuration deviation in the disinfection response parameter model, and the automatic update logic is triggered.

[0014] In the above technical solution, the technical effects and advantages provided by the present invention are: 1. This invention achieves multi-parameter dynamic optimization of the fabric cleaning and disinfection process by constructing a disinfection response parameter model based on pollution source information, combined with real-time monitoring and feedback control mechanisms. The system not only intelligently sets wash temperature, time, and disinfectant concentration based on fabric material and target bacterial species, but also automatically senses temperature fluctuations and changes in chemical concentration during operation, adjusting the washing strategy in real time to ensure effective disinfection while effectively protecting fabrics from damage.

[0015] 2. This invention incorporates a post-wash residual bacteria analysis and model self-learning mechanism, enabling feature extraction and comprehensive analysis of dynamic temperature deviations and actual inhibition results, optimizing subsequent operating parameters accordingly. This closed-loop update strategy significantly improves the system's adaptability and stability against multidrug-resistant bacteria, offering significant technical advantages such as precise disinfection, strong fabric compatibility, and optimized energy efficiency control. It is particularly suitable for use in healthcare, elderly care, public health, and other sectors with high hygiene control requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0017] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0019] Example 1, please refer to Figure 1 As shown, the operating method of the fabric cleaning and disinfecting machine for controlling multidrug-resistant bacteria described in this embodiment includes: S100: Obtain the pollution intensity level, fabric material type, and target bacterial species spectrum of the target fabric, and construct a pollution source information set P; S200: Based on the pollution source information set P, a preset disinfection response parameter model is called to determine the washing temperature T, washing time t, disinfectant type and concentration parameter C; S300: Based on the fabric type and its temperature and chemical resistance, a matching washing program set W is retrieved from the washing machine database and dynamically adjusted according to the disinfection requirements; S400: During the main washing process, the water temperature and detergent concentration are monitored in real time, and the washing parameters are adjusted according to the preset feedback control model; S500: extract and analyze the dynamic temperature deviation value of the washing stage during the washing and disinfection process and the critical inhibition and killing threshold of the target bacteria in the residual bacteria detection data, and automatically update the disinfection response parameter model of the fabric washing according to the analysis results.

[0020] In this method, information extraction and contamination signature identification are first performed on the target fabrics entering the cleaning process to construct a contamination source information set for subsequent intelligent operation control. This step, a critical foundational step before the disinfection process, directly influences the scheduling strategy for subsequent cleaning and disinfection parameters, the matching accuracy of disinfectants, and the efficiency of equipment resource control.

[0021] Specifically, the pollution source information set includes the following three key attribute information: fabric pollution intensity level, fabric material type, and target bacterial species spectrum information. The acquisition method and processing flow of each type of information are as follows: The fabric contamination intensity grade is mainly used to characterize the microbial load level and organic stain content on the fabric surface and inside the fibers, and is an important indicator of the initial hygiene status of the fabric. Its assessment method can be based on a combination of the following three methods: Source identification method: This method makes a prediction based on information about the fabric's location and lifecycle. For example, fabrics from intensive care units, isolation wards, or infectious disease departments are assigned a high contamination intensity level by default; fabrics from general departments or nursing facilities are assigned a medium level; and fabrics from general household clothing or dormitory areas are assigned a low level.

[0022] Sensor analysis: Before washing, the machine uses built-in spectral, image, or odor recognition sensor modules to perform an initial screening of fabrics for contaminants. For example, the intensity of signals such as fluorescence reactions and organic protein residues is measured, and the contamination intensity is automatically determined using a pre-set threshold model within the machine.

[0023] Manual labeling and learning method: The operator selects the pollution level (such as mild, moderate, and severe) through the interface selection operation before putting the fabric in, and combines the previous washing residual bacteria monitoring results to conduct model training to form a dynamic adjustment system for the pollution level.

[0024] The pollution intensity level can be divided into 3 to 5 levels through the above method. The system uses graded numbers inside, for example, pollution level P∈{1, 2, 3}, corresponding to light, moderate and heavy pollution respectively.

[0025] Fabric material identification is mainly used to determine its temperature resistance, chemical corrosion resistance, mechanical tensile strength, etc., to avoid damage in high temperature or strong chemical environments. Fabric material identification can be done in the following ways: RFID code identification: For fabrics embedded with electronic tags (such as RFID), the fabric material type is obtained by scanning and reading the code information, such as pure cotton, polyester, nylon, medical non-woven fabrics, etc.

[0026] Image texture and fiber analysis method: Use the device's built-in camera system or microscope sampling device to extract the texture of the fabric fiber image, and use a deep learning model to classify the image to determine the material type.

[0027] Manual input and system memory method: When the device recognizes an unknown fabric for the first time, the operator can manually select the fabric type. The system will save the image and parameters of this type of fabric and automatically compare them for subsequent recognition.

[0028] Different material types are represented by codes, such as cotton as F1, polyester as F2, non-woven fabric as F3, wool as F4, etc. The system associates material properties with washing constraints in the form of a matrix to limit temperature, type of disinfectant and mechanical program.

[0029] The types of multidrug-resistant bacteria have a key impact on the selection and duration of disinfectant action, so it is necessary to determine the target bacterial species spectrum as much as possible before operation. This information can be obtained in the following ways: Using a site risk profile matching method, high-risk sites are pre-set with corresponding bacterial species risk models, such as infectious disease departments associated with MRSA and CRE, burn wards associated with Pseudomonas aeruginosa, and long-term care facilities associated with Acinetobacter baumannii. The system matches the relevant bacterial species profile based on fabric source tags.

[0030] Retrospective method for detecting historical washing residual bacteria: The equipment is equipped with a periodic sampling and culture device, which combines the previous residual bacteria detection data to infer the dominant bacterial group that the current batch may carry.

[0031] The bacterial species information is modeled using spectrum numbering, such as G1 for Staphylococcus aureus, G2 for Pseudomonas aeruginosa, and G3 for Acinetobacter baumannii. The system records the target bacterial spectrum G = {G1, G3} in a set manner.

[0032] After obtaining the above three types of information, the pollution level code P, fabric material type code F, and target bacterial species spectrum G are combined to form the pollution source information set InfoSet = {P, F, G}. The system uses this information set as input and passes it to the disinfection response parameter generation module, which serves as the core basis for parameter scheduling.

[0033] For example, in an actual operation case, a batch of fabrics comes from an infectious disease ward and is used frequently, so it is marked as pollution level P=3. The material is medical non-woven fabric, which is sensitive to temperature and is recorded as F=F3. It is known that there is a risk of infection by Acinetobacter baumannii and Pseudomonas aeruginosa in the ward, corresponding to G = {G2, G3}. The pollution source information set is: InfoSet = {3, F3, {G2,G3}}.

[0034] After receiving this information set, the system will automatically disable the high-temperature sterilization program and instead adopt a low-temperature, long-acting disinfection program using peracetic acid or a compound quaternary ammonium salt. At the same time, it will extend the action time and set multiple soaking steps in the program to increase the killing rate of drug-resistant bacteria and avoid damage to the fabric.

[0035] In summary, through the systematic identification and fusion modeling of pollution level, fabric material and bacterial species spectrum, precise and differentiated control of the fabric cleaning and disinfection process can be achieved, which is one of the basic technical links of the present invention.

[0036] In the present invention, after the contamination source information set (denoted as P) is constructed, the fabric cleaning and disinfection response phase begins. The core of this phase is to automatically determine the optimal washing and disinfection process configuration parameters based on the various input parameters of the contamination source information set by invoking a preset disinfection response parameter model. These parameters include, but are not limited to, the washing temperature T, the washing time t, and the disinfectant type and concentration parameter C.

[0037] The disinfection response parameter model described in the present invention is a dynamic matching and optimization model under multivariable conditions, with the following three core functions: After inputting the pollution source information set P, the cleaning configuration parameter set (T, t, C) can be quickly output; Ability to make multi-objective trade-offs based on microbial characteristics, fabric tolerance and pollution level; It has the ability to continuously learn and dynamically update to improve long-term adaptability and model accuracy.

[0038] The model construction in the present invention takes the pollution source information set P as input and outputs the parameter set (T, t, C), namely: Input P = {P_level, F_type, G_species}, where: P_level represents the pollution intensity level, ranging from 1 to 5; F_type indicates the type of fabric material, such as cotton, polyester, non-woven fabric, etc.; G_species represents the target microbial spectrum, which can be a single species or a collection of multiple multidrug-resistant bacteria; Output parameter set: T represents the recommended washing temperature in degrees Celsius; t represents the washing and disinfection time in minutes; C represents the type of disinfectant selected and its recommended concentration, expressed in mass volume percentage (w / v%) or mass concentration (mg / L).

[0039] This model is constructed by combining rule mapping with multidimensional regression to achieve parameter conversion from P →(T, t, C).

[0040] First, we collected a large amount of real-world cleaning case data. This data included contamination levels, fabric types, bacterial strains, temperatures and times used, disinfectant types and concentrations, and residual bacterial test results (e.g., colony count reduction ratio, target bacteria detection rate, etc.) after each cleaning batch. Based on this data, we constructed a sample set consisting of input variables and target output variables.

[0041] The construction goal is to avoid damage to the fabric material while meeting the sterilization effect (for example, the target bacteria killing rate is ≥99.99%), and to control the overall energy and resource consumption.

[0042] First, a set of basic rule mapping tables is established. The rules are based on expert experience, industry standards and existing scientific research data to define the basic parameter configuration ranges under different pollution levels and bacterial types. For example: For cotton fabrics with a contamination level of 3 and target bacteria MRSA, the recommended temperature is 60–70°C, the exposure time is 15–30 minutes, and the concentration of quaternary ammonium salt disinfectant is 500–1000 mg / L; For non-woven fabrics with a contamination level of 5 and a combination of Pseudomonas aeruginosa and Acinetobacter baumannii, the recommended temperature is no higher than 50°C, the exposure time needs to be extended to 40–60 minutes, and a low-temperature broad-spectrum disinfectant containing peracetic acid should be used at a concentration of 1000–2000 mg / L.

[0043] This set of rules forms the initial matching matrix, which is used to constrain the model parameter space and provide the initial fitting surface.

[0044] On the basis of rule mapping, the weighted multivariate regression analysis method is further used to establish a continuous response relationship function, that is, the model is defined as a multi-input single-output or multi-output mapping problem.

[0045] Let pollution level P_level be the input variable ; Fabric material type F_type is mapped as a categorical variable to a numerical variable ; The species spectrum G_species is converted into multiple input dimensions through one-hot encoding to , represents the tolerance score of each bacterial species; Let T, t and C be the objective functions respectively 、 、 ; Finally, three sets of fitting functions are constructed: → Used to predict temperature; → Used to predict time; → Used to recommend disinfectant type and concentration.

[0046] The above function uses sample training and residual optimization to obtain the optimal fitting parameters under the minimum squared error. Regularized regression is used to control model overfitting, and a cross-validation strategy is introduced to improve generalization ability.

[0047] To ensure that the model output parameters are fault-tolerant to fabric safety and sterilization reliability, a tolerance interval is introduced in each output dimension. For example: The temperature output T can fluctuate up and down by ±3°C; Time output t can float ±10%; The disinfectant concentration C may deviate from the permitted concentration by ±15%.

[0048] If the bacterial killing score corresponding to a certain output configuration does not reach the set threshold (for example, the sterilization score is <95%), the model will provide feedback correction to the point and update the fitting surface using neighborhood interpolation.

[0049] In addition, the model assigns different weights to the input parameters. By default, the bacterial species spectrum has the greatest impact on the output, followed by the pollution level, and finally the fabric type. The weight ratio is as follows: G_species (bacterial species spectrum): 0.5; P_level (pollution level): 0.3; F_type (fabric type): 0.2; This weight factor can be dynamically updated through actual application feedback to achieve personalized optimization of the model.

[0050] The response model in the present invention has a continuous learning function. After each run, the system collects the residual bacteria information of the washing batch and the fabric integrity evaluation data, re-evaluates the current parameter configuration effect, and uses the gradient descent strategy to fine-tune the model parameters to enhance the model's adaptive ability.

[0051] The model update follows the principle of “stable output and local iteration”, that is, corrections are only made to local areas with large prediction errors without interfering with the overall fitting structure, thereby ensuring that the stability and sensitivity of the response model are taken into account.

[0052] After the model completes the calculation, the system will output specific operating parameters: The recommended washing temperature is T (e.g. 62°C); The recommended action time is t (e.g. 28 minutes); The recommended disinfectant is a certain type of compound quaternary ammonium salt solution with a recommended concentration of 850 mg / L.

[0053] The control system automatically configures the equipment to execute programs based on these outputs, completing safe and efficient disinfection of different types of fabrics.

[0054] In this invention, after acquiring the contamination source information set and determining the wash temperature T, exposure time t, disinfectant type, and concentration parameter C using the disinfection response parameter model, the system enters the cleaning process program allocation phase. The key tasks of this phase are: based on the fabric type and its temperature and chemical resistance, searching for a matching set of cleaning programs W from the washing machine database, and dynamically adjusting this set of programs based on current disinfection requirements to meet the dual requirements of sterilization and fabric protection.

[0055] This process is divided into two main stages: The following describes the two processes of retrieving and initializing the standard program set W and dynamically adjusting and adapting the program set based on disinfection requirements.

[0056] The retrieval and initialization of the standard cleaning program set W is as follows: The laundry equipment is pre-set with multiple standard cleaning programs, which are stored in a local or remote database. Each program consists of a series of parameters, including but not limited to: Initial water inlet temperature and heating rate; The total duration and segment structure of the main wash and pre-wash; Detergent or disinfectant injection time point; stirring intensity and rhythm; frequency and intensity of dehydration phases; Number of flush cycles; Temperature upper limit and mechanical action constraints.

[0057] Each program is tied to a specific fabric material type and function tag, such as "Cotton High-Temperature Standard Wash," "Synthetic Fiber Sensitive Mode," "Medical Fabric Sterilization Program," and "Heavy Pollution and Acid-Resistant Program." Program metadata is stored in a structured format, facilitating conditional filtering and matching.

[0058] When the system obtains the pollution source information set P, especially the fabric material type F (such as pure cotton, polyester, non-woven fabric, wool, etc.), it first preliminarily screens the available program set WO through material index matching to construct a standard program candidate set.

[0059] For example: if the fabric type is F = F1 (pure cotton), the system will filter all programs marked as "F1 compatible" from the database to form the set WO = {W1, W2, W4, W6}; If the fabric type is F = F3 (non-woven), only select programmes with low temperature protection and gentle mechanical action, e.g. WO = {W7, W9}.

[0060] This step ensures that the selected program is safe for the fabric by default and does not damage the material due to high temperatures or mechanical stress.

[0061] Dynamic adjustment of the program set W based on disinfection requirements, specifically: After the standard program set W0 is established, it needs to be dynamically adjusted according to the specific disinfection requirements of this operation to improve the killing ability of target multidrug-resistant bacteria. In this stage, the following five-step strategy is used to implement dynamic program configuration: The system extracts the corresponding disinfection level requirements based on G (target bacterial species spectrum) in the pollution source information set and the output results (T, t, C) of the response parameter model. For example: If the target bacteria is Pseudomonas aeruginosa, the disinfection requirement level is set to D = high; If the target bacteria is a single non-spore-forming bacteria, the disinfection requirement level is D = medium; If bacterial spores (such as Clostridium difficile) are involved, D = very high.

[0062] The system uses D as the primary driving variable of the dynamic adjustment factor to control subsequent strategies.

[0063] The heating rate and maximum temperature set at the initial stage of the program will be adjusted according to the T value: If the model outputs a recommended temperature T′ that is higher than the upper temperature limit of the original program, the system will perform a temperature curve reset operation; The temperature adjustment control formula is: New temperature target = max{original program temperature, model recommended temperature T′}, but it must not exceed the upper temperature limit of the fabric ;in This is derived from a table of fabric properties, e.g. 55°C for non-woven fabrics and 90°C for cotton. If the limit is exceeded, the priority is to maintain fabric safety and compensate by extending the disinfectant action time in subsequent stages.

[0064] Based on the model's output of the action time t', the main wash and disinfection soak phases of the program are stretched. The system inserts a "high-concentration rest period" or "low-speed stirring and heat preservation period" into the wash program structure to ensure the disinfectant continues to act in high-load areas.

[0065] The duration adjustment logic is as follows: If t′ is longer than the total duration of the current program, a delay is inserted into the main wash section or the middle section; If t′ is shorter than the original program duration, equivalent sterilization conversion is performed based on the recommended concentration of the disinfectant, and the time is allowed to be shortened.

[0066] The present invention supports flexible disinfectant injection timing and multiple flushing combinations: If a high-concentration, low-temperature disinfectant is used, the injection point should be set at the low-speed stirring section; If volatile oxidants (such as peracetic acid) are used, the exposure period should be shortened and rinsed quickly after injection to prevent corrosion of the fabric.

[0067] The system adjusts the three parameters of disinfectant injection time point T_inject, retention time t_hold, and rinse times n_rinse in a coordinated manner to form the most effective disinfectant management path for the target strain.

[0068] After completing these adjustments, the system merges the original standard program structure with the dynamic adjustment factors to form a new customized program structure, W′: W′ = W_i + ΔT + Δt + ΔC + injection strategy + rinse strategy. ΔT represents the temperature curve adjustment, Δt is the duration compensation value, and ΔC is the disinfectant concentration adjustment factor. All adjustments are written by the controller to the washing machine's control chip, creating a traceable cleaning record for the current batch.

[0069] To prevent system operation imbalance caused by multiple parameter adjustments, the present invention also designs the following safety control mechanism: When the disinfection goal conflicts with fabric protection, the system will choose a non-destructive solution for the fabric and make alternative compensations by increasing the dosage, extending the time, etc. During the dynamic adjustment process, the system will simultaneously evaluate changes in resource consumption such as water, electricity, and detergent to avoid exceeding the resource limit for a single wash. All parameter adjustments will form an operation log, which will serve as the data basis for subsequent model updates and controller firmware upgrades.

[0070] In the present invention, during the execution of the main washing program, the system not only executes the control process according to the static set values, but also monitors key operating parameters such as water temperature, water quality, conductivity, detergent concentration, etc. in real time through the embedded sensor system, and dynamically fine-tunes the key washing process parameters through the feedback control model to improve the disinfection effect of multi-drug resistant bacteria control and ensure the physical integrity of the fabric material and the energy efficiency of the system operation are optimized.

[0071] The core of this process lies in building a feedback control model that integrates real-time data collected during the washing process with preset expectations, dynamically correcting system operating parameters and achieving multivariable closed-loop control. The following details the feedback model's construction principles and adjustment mechanisms.

[0072] This paper uses an improved fuzzy-proportional-integral-differential (Fuzzy-PID) control model to construct a feedback control system for washing machines. This model combines the real-time error adjustment capabilities of a traditional PID controller with the multi-objective decision logic of a fuzzy controller, making it suitable for dynamic adjustment of multiple parameters in nonlinear systems.

[0073] The control target variables include: Water temperature T(t): current water temperature; Detergent concentration C(t): mass concentration of the active agent in the solution; Disinfectant action time : The cumulative duration of the active ingredient's action at the current stage; Stirring intensity S(t): motor drive frequency or torque index; The system expected value comes from the output of the response parameter model established in the previous step and is set as the target value: Desired temperature: ; Desired concentration: ; Expected duration of action: ; During the operation, the system collects the actual values ​​T(t), C(t), , and then calculate the deviation after comparing with the expected value , , .

[0074] The system predefines several fuzzy language sets (such as "low temperature", "rapid concentration drop", "dramatic temperature fluctuation", etc.), and establishes membership functions to map actual deviation values ​​such as ΔT and ΔC into fuzzy levels.

[0075] For example: When , the corresponding membership degree of “low temperature” is 0.85; When , the corresponding membership degree of “concentration decreases rapidly” is 0.90.

[0076] Based on these fuzzy inputs, the system uses a fuzzy inference rule base to determine the direction and strength of the control output.

[0077] Some examples of rules are as follows: IF the temperature is low AND the concentration is normal THEN increase the heating power (medium); IF concentration drops rapidly AND temperature is normal THEN delay flushing; IF THE TEMPERATURE FLUCTUATIONS ARE EXTREME THEN SLOW DOWN THE HEATING RATE.

[0078] Fuzzy rules are used to determine the direction of the control strategy, while the fine-tuning of the amplitude is completed by the PID controller.

[0079] The PID controller uses a three-term error superposition model, namely: ; Where: e(t) is the current deviation value (such as ΔT); They are the proportional, integral, and differential coefficients, respectively, which are set according to experience and measured parameters; U(t) is the control output, such as the heater power adjustment percentage or the detergent pump speed correction amount.

[0080] The output level of the fuzzy logic controller will directly affect the dynamic adjustment of the coefficients in the PID controller, thereby achieving fuzzy-PID linkage response.

[0081] After the feedback control model is established, the following parameters are mainly adjusted dynamically during actual operation: The system collects water temperature T(t) in a period of 1 to 5 seconds and compares it with the target value. If the temperature deviation ΔT exceeds the set tolerance (e.g. ±1.5°C), the temperature control loop is immediately triggered.

[0082] ΔT>+1.5°C (too high temperature): the system reduces the heater power or performs cold water replenishment when necessary; (Low temperature): The system increases heater output and reduces drum agitation frequency to reduce heat loss.

[0083] Temperature control can maintain the water temperature stable Within.

[0084] The detergent concentration C(t) is detected by the conductivity and color sensors. If the deviation ΔC>100 mg / L, the system will perform a micro-pulse delivery on the detergent pump.

[0085] During the main wash phase, the dose can be adjusted every 3 minutes, and the maximum single correction dose shall not exceed 5% of the total dose; If the concentration downward trend continues, the system will start the "delayed flushing" program and postpone the start time of the next flushing section to extend the effective contact time.

[0086] System based Calculate whether the current cumulative disinfection time has reached the threshold. If the efficiency is reduced due to low temperature or insufficient concentration, the system will: Automatically extend the running time of the current cleaning stage; or insert an additional buffer segment to achieve time compensation to ensure that the total action time meets the model recommended value.

[0087] By reading the motor feedback frequency and torque value, the drum load status is judged, and the stirring rhythm is adjusted comprehensively based on the temperature and concentration parameters. For example: When the temperature is low but the concentration is high, moderately increase the stirring frequency to enhance the physical effect; When the temperature is high and the concentration is low, reduce the stirring intensity to avoid excessive release of chemical activity.

[0088] The system can select the stirring mode (such as forward and reverse alternation, high-frequency impact or low-speed suspension, etc.) based on the bacterial strain resistance label to optimize the sterilization effect.

[0089] The feedback control model in the present invention supports adaptive updating. After completing a washing cycle, the system automatically records the key parameter change trajectory, control response log and residual bacteria information after cleaning to form an operation record set.

[0090] Through "deviation attribution analysis" and "control response learning", the system can iteratively correct the following: Adjust the membership function intervals in the fuzzy controller; Optimize PID coefficient to adapt to different fabrics or bacterial species; Identify the impact of external factors such as ambient temperature and water hardness on system stability, and improve the robustness of full-cycle control.

[0091] In this invention, in order to improve the control effect of multidrug-resistant bacteria during the fabric cleaning process, the system will collect and analyze key dynamic data during the operation after the washing and disinfection process is completed, especially: Dynamic temperature deviation value in the washing stage, and critical inhibition threshold of target bacteria in residual bacteria detection data.

[0092] By jointly analyzing the above two characteristic parameters from different dimensions, the sterilization effect of the parameter configuration in this operation can be quantified, thereby realizing automatic updating and self-learning optimization of the disinfection response parameter model.

[0093] The steps for extracting and calculating the dynamic temperature deviation value include: During the main washing process, the system obtains the current water temperature data through the built-in temperature sensor at a fixed sampling period (for example, every 5 seconds) and records it to form a time series: ;in, to represents the time node from the start of washing to the completion of disinfection, and n represents the total number of time points. The system also records the target constant temperature value output by the disinfection response parameter model during this operation, which is recorded as .

[0094] The system is based on the actual temperature of each sampling point With target temperature The difference between them is used to calculate the instantaneous temperature deviation sequence: ; Then, the deviation of the whole process is normalized and averaged to obtain the dynamic temperature deviation value ΔT_avg of this washing stage, which is recorded as: This is a quantitative evaluation indicator of the actual temperature control stability. A smaller ΔT_avg indicates higher temperature control accuracy; a larger ΔT_avg indicates frequent fluctuations or control lag, which may affect sterilization efficiency.

[0095] The steps for extracting and calculating the critical inhibition and killing threshold of the target bacterial species are as follows: After the fabric is washed and dried, the system collects the residual bacterial data after this operation and uses biological detection modules (such as rapid culture method, fluorescent staining counting, or optional PCR rapid detection module) to identify the survival status of the target bacteria (from the contamination source information set G) in the residual sample after washing.

[0096] Set a target bacteria before running The initial bacterial load is: (Unit: CFU / mL); After washing, the residual load of the bacteria was detected as follows: ; This operation is for The actual killing rate is defined as: ; For example: If the initial quantity is 10 6 CFU / mL, and the residual after washing is 10² CFU / mL, then the inhibition rate is: ; The system pre-sets recommended inhibition and killing thresholds for each type of common multi-drug resistant bacteria, called: For example, the threshold for MRSA is ≥ 99.99%; for Pseudomonas aeruginosa, ≥ 99.999%; for Acinetobacter baumannii, ≥ 99.9%. The actual inhibition rate is compared with the recommended threshold to form a difference , the expression is: ; If the difference is negative, it means that the current program does not achieve the recommended sterilization effect; if it is positive, it means redundant disinfection and there is a problem of excessive resources.

[0097] For all target bacterial species Perform weighted average and combine the weight of each strain’s hazard level , and the critical inhibition and killing threshold of the target bacteria was obtained : .

[0098] The dynamic temperature deviation value and the critical inhibition threshold are converted into a comprehensive feature vector, and the comprehensive feature vector is used as the input of the machine learning model. The machine learning model uses each set of comprehensive feature vectors to predict the error score value label of the disinfection response parameter model for the current operation control performance as the prediction target, and takes minimizing the sum of the prediction errors of the error score value labels of all disinfection response parameter models for the current operation control performance as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The error score value of the disinfection response parameter model for the current operation control performance is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0099] When the error score E_model of the disinfection response parameter model for the control performance of this operation exceeds the set threshold E_thresh (such as 5.0), the system determines that there is a configuration deviation in the current model and triggers the automatic update logic.

[0100] The system will record the operation of this batch (including etc.) as new training samples and update the following model modules: Adjust the weight coefficient in the temperature response function to improve the temperature control sensitivity; Lower or increase the recommended disinfectant concentration C' or action time t' for the corresponding bacterial species; Refit the optimal temperature-concentration combination interval for certain specific bacterial species.

[0101] Using gradient descent or incremental learning, only local areas of the model are corrected, preserving the overall stability of the model. The new model version is automatically archived, overwriting the original parameter table.

[0102] After each correction operation is completed, the system records: Original model parameters; Actual operation feedback; Correct the logic path; New version number and activation time.

[0103] All historical model data supports retrospective analysis and fault comparison, ensuring the traceability and stability of the model evolution process.

[0104] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A method for operating a fabric cleaning and disinfecting machine for controlling multidrug-resistant bacteria, characterized in that: Including: S100: obtain the pollution intensity level, fabric material type and target bacterial species spectrum information of the target fabric, and construct the pollution source information set P; S2 00: Based on the pollution source information set P, call the preset disinfection response parameter model to determine the washing temperature T, washing time t, disinfectant type and concentration parameter C; S3 00: Based on the fabric type and its temperature and chemical resistance, the matching washing program set W is retrieved from the washing machine database and dynamically adjusted according to the disinfection requirements; S400: During the main washing process, the water temperature and detergent concentration changes are monitored in real time, and the washing parameters are adjusted according to the preset feedback control model; S5 00: Extract and analyze the dynamic temperature deviation value of the washing stage during the washing and disinfection process and the critical inhibition and killing threshold of the target bacteria in the residual bacteria detection data, and automatically update the disinfection response parameter model of fabric washing based on the analysis results.

2. The method for operating a fabric cleaning and disinfecting machine for controlling multidrug-resistant bacteria according to claim 1, characterized in that: In S200, the disinfection response parameter model is a multivariate fitting model constructed based on a pollution source information set, wherein the pollution source information set includes pollution intensity level, fabric material type, and target bacterial species spectrum information. The model is constructed by the following steps: The historical fabric cleaning sample data was collated, and the pollution level, material type, and bacterial species information corresponding to each batch were extracted as input variables; Extract the actual operating parameters of the corresponding batch, including washing temperature, action time, and disinfectant type and concentration, as the target output variables; Perform multidimensional regression modeling on input and output data to construct washing temperature prediction function, time prediction function and disinfectant concentration recommendation function; Introducing tolerance adjustment mechanisms and weight factors into the modeling process to control the acceptable floating range of temperature, time, and disinfectant concentration, as well as the weighted impact of each input variable; The model results are cross-validated and corrected through error feedback to achieve dynamic adaptation of disinfection response parameters.

3. The method for operating a fabric cleaning and disinfecting machine for controlling multi-drug resistant bacteria according to claim 2, characterized in that: In S300, the dynamic adjustment according to the disinfection requirements includes: Extracting the target washing temperature, action time, and disinfectant type and concentration output by the disinfection response parameter model; Screening the standard programs that match the fabric material type in the preset cleaning program set to form an initial candidate program set; Adjust the temperature rise curve, disinfection section holding time, and disinfectant injection timing in the candidate program according to the disinfection requirement level of the target bacterial species; Increase the program setting temperature without exceeding the upper temperature limit of the fabric; Dynamically adjust the number and rhythm of flushing sections to meet the requirements of different types of disinfectants for residual suppression; The adjusted program parameters are combined to form the final operating program for the execution of this fabric cleaning and disinfection task.

4. The method for operating a fabric cleaning and disinfecting machine for controlling multidrug-resistant bacteria according to claim 1, characterized in that: In S400, adjusting the washing parameters according to the preset feedback control model includes: A composite feedback control model comprising fuzzy control and proportional-integral-differential control is constructed, wherein the model uses water temperature, detergent concentration, and disinfection time acquired in real time during the washing process as input variables; Set target values ​​for washing parameters, including target water temperature, target detergent concentration, and desired disinfection time; Calculate the deviation between the current parameter and the target value, and determine the parameter deviation level through the fuzzy reasoning module; Based on the deviation level, the corresponding adjustment command direction is output and input into the proportional integral derivative controller to generate a continuous control signal; The control signal is used to adjust the heater output, detergent delivery rate, stirring frequency or washing phase time to dynamically correct deviations.

5. The method for operating a fabric cleaning and disinfecting machine for controlling multidrug-resistant bacteria according to claim 1, characterized in that: In S500, the calculation step of the dynamic temperature deviation value includes: During the main washing process, the current water temperature data is obtained at a fixed sampling period and recorded to form a time series: ;in, to represents the time node from the start of washing to the completion of disinfection, and n represents the total number of time points; at the same time, the target constant temperature value output by the disinfection response parameter model during this operation is recorded, which is recorded as ; Based on the actual temperature at each sampling point With target temperature The difference between the two is used to calculate the instantaneous temperature deviation series : ; Then, the deviation of the whole process is normalized and averaged to obtain the dynamic temperature deviation value ΔT_avg of this washing stage, which is recorded as: .

6. The method for operating a fabric cleaning and disinfecting machine for controlling multi-drug resistant bacteria according to claim 5, characterized in that: The calculation steps of the critical inhibition and killing threshold of the target bacterial species are as follows: After the fabric is washed and dried, the residual bacterial data after this run is collected and the survival status of the target bacteria species in the residual sample after washing is identified; Set a target bacteria before running The initial bacterial load was ; After washing, the residual load of the bacteria was detected to be ; This operation is for The actual killing rate is defined as: ; Set recommended inhibition and killing thresholds for each type of common multidrug-resistant bacteria in advance Each target bacteria The actual inhibition rate is compared with the recommended threshold to form a difference , the expression is: ; For all target bacterial species The critical inhibition and killing threshold of the target bacterial species was obtained by taking a weighted average and combining the weight of the hazard level of each bacterial species.

7. The method for operating a fabric cleaning and disinfecting machine for controlling multi-drug resistant bacteria according to claim 6, characterized in that: The dynamic temperature deviation value and the critical inhibition threshold are converted into a comprehensive feature vector, and the comprehensive feature vector is used as the input of the machine learning model. The machine learning model uses each set of comprehensive feature vectors to predict the error score value label of the disinfection response parameter model for the current operation control performance as the prediction target, and takes minimizing the sum of the prediction errors of the error score value labels of all disinfection response parameter models for the current operation control performance as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The error score value of the disinfection response parameter model for the current operation control performance is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

8. The method for operating a fabric cleaning and disinfecting machine for controlling multi-drug resistant bacteria according to claim 7, characterized in that: When the error score E_model of the disinfection response parameter model for the control performance of this operation exceeds the set threshold E_thresh, it is determined that there is a configuration deviation in the disinfection response parameter model, and the automatic update logic is triggered.

Images