A method and system for assessing anal cleansing effectiveness

Abstract

The application belongs to the technical field of biological simulation, and discloses a method and system for evaluating the cleaning effect of an anus. The method controls an execution assembly to wipe the anus anatomical structure according to a target wiping strategy, and controls a multi-modal sensing assembly to collect target parameters on the anus anatomical structure to obtain target monitoring data. According to the residual dirt amount in the target monitoring data, a cleaning efficiency score is calculated, and a target mechanical simulation model is called to calculate a skin irritation risk score according to the humidity penetration gradient value and the pressure value in the target monitoring data and the test sample, so as to form a cleaning effect evaluation report with the cleaning efficiency score, thereby realizing simulation and prediction of the anus cleaning process, greatly improving the simulation test effect and accuracy, and providing the cleaning effect evaluation report to indicate the safety of the test sample.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of biological simulation, and particularly relates to an anus cleaning effect evaluation method and system. BACKGROUND

[0002] Cleaning of the anus is crucial for human health, and reliable testing methods are needed in the performance evaluation of wet toilet paper products. However, the existing anus cleaning effect testing technology has technical problems such as poor ethics, insufficient simulation, and inaccurate detection methods, which cannot meet the growing testing needs of sanitary product research and development.

[0003] Therefore, the prior art still needs to be improved and improved. SUMMARY

[0004] The main purpose of the present application is to provide an anus cleaning effect evaluation method and system, which aims to solve the problem of poor test effect and low accuracy in simulating the test of the anus cleaning effect in the prior art.

[0005] In order to achieve the above purpose, the following technical solutions are adopted in the present application:

[0006] In a first aspect, the present application provides an anus cleaning effect evaluation method, comprising:

[0007] The control execution component wipes the anus anatomical structure according to the target wiping strategy of the to-be-tested sample, and controls the multi-modal sensing component to collect target parameters on the anus anatomical structure to obtain target monitoring data;

[0008] According to the residual dirt amount in the target monitoring data, a cleaning efficiency score is calculated, and a target mechanical simulation model is called to calculate a skin irritation risk score according to the humidity penetration gradient value and the pressure value in the target monitoring data and the to-be-tested sample;

[0009] After the skin irritation risk score is graded according to a preset grade, the skin irritation risk score and the cleaning efficiency score constitute a cleaning effect evaluation report of the to-be-tested sample.

[0010] In some embodiments, the control execution component wipes the anus anatomical structure according to the target wiping strategy of the to-be-tested sample, and the method further comprises:

[0011] The execution component is controlled to simulate wiping the anus anatomical structure according to a preset wiping strategy, and the multi-modal sensing component is controlled to collect target parameters on the anus anatomical structure to obtain simulation monitoring data;

[0012] The initial mechanical simulation model is called to perform finite element simulation based on initial model parameters to obtain comprehensive response data, which is compared with the simulation monitoring data to obtain calibration parameters.

[0013] The skin material parameters in the initial model parameters are adjusted according to the calibration parameters to obtain the target mechanical simulation model.

[0014] In some embodiments, after the initial mechanical simulation model is called to perform finite element simulation based on initial model parameters to obtain comprehensive response data, the calibration parameters are obtained by comparing the simulation monitoring data.

[0015] The initial mechanical simulation model is called to perform multiple forward finite element simulations based on initial model parameters until the number of iterations reaches a preset maximum number of iterations to obtain predicted friction and predicted equivalent liquid permeability.

[0016] The actual friction is analyzed according to the pressure distribution change in the simulation monitoring data and the motion parameters of the execution assembly, and compared with the predicted friction to obtain the baseline coefficient of friction and elastic modulus.

[0017] The skin layer actual equivalent liquid permeability is obtained by analyzing the liquid penetration speed, distribution and moisture change data of the simulation monitoring data at different depths of the skin, and compared with the predicted equivalent liquid permeability to obtain the calibration coefficient of the equivalent liquid permeability.

[0018] The motion parameters include: speed, acceleration, joint torque, position and motion direction; the target parameters include: pressure value, residual dirt amount and humidity penetration gradient value; the skin material parameters include: equivalent liquid permeability, elastic modulus and friction coefficient of the skin layer.

[0019] In some embodiments, before the control execution assembly wipes the anal anatomical structure according to the target wiping strategy, it further includes:

[0020] The target mechanical simulation model is called to analyze the simulation monitoring data collected under different wiping strategies to obtain a cleaning effect evaluation value.

[0021] The reinforcement learning model is called to perform deep learning on the cleaning effect evaluation value and the state space in the simulation monitoring data to obtain a reward value of different wiping strategies.

[0022] The policy parameters of the reward function of the reinforcement learning model are updated according to the reward value to obtain the target wiping strategy.

[0023] In some embodiments, the cleaning efficiency score is calculated according to the residual soil amount in the target monitoring data, and a target mechanical simulation model is called to calculate a skin irritation risk score according to a humidity penetration gradient value and a pressure value in the target monitoring data, and the sample to be tested, including:

[0024] The cleaning efficiency score is calculated after the depth removal rate, the soil removal rate, and the cleaning uniformity size are calculated according to the residual soil amount on the radial fold layer of the anal anatomical structure in the target monitoring data;

[0025] The target mechanical simulation model is called to analyze the skin moisture balance and the chemical irritation size according to the humidity penetration gradient value and the component information of the sample to be tested, and to evaluate the mechanical irritation size according to the pressure value in the target monitoring data, so as to calculate the skin irritation risk score according to the skin moisture balance, the chemical irritation size, and the mechanical irritation size.

[0026] In some embodiments, the depth removal rate, the soil removal rate, and the cleaning uniformity size are calculated according to the residual soil amount on the radial fold layer of the anal anatomical structure in the target monitoring data, including:

[0027] The depth removal rate is calculated by comparing the initial soil amount at the bottom of the fold of the radial fold layer with the residual soil amount at the corresponding position in the target monitoring data;

[0028] The soil removal rate is calculated by comparing the initial total soil amount on the radial fold layer with the residual total soil amount at the corresponding position;

[0029] The standard deviation size between the residual soil amounts of different target regions in the radial fold layer is calculated to obtain the cleaning uniformity size;

[0030] The target mechanical simulation model is called to analyze the skin moisture balance and the chemical irritation size according to the humidity penetration gradient value in the target monitoring data and the component information of the sample to be tested, and to evaluate the mechanical irritation size according to the pressure value in the target monitoring data, including:

[0031] The transdermal water loss rate and the skin water content change are calculated according to the real-time change of the humidity penetration gradient value to analyze the skin moisture balance;

[0032] The chemical irritation size is predicted according to the humidity penetration gradient value, and the pH value and the surfactant of the sample to be tested;

[0033] According to real-time changes of the pressure values, stress distribution of the surface and the interior of the radial fold layer is calculated to evaluate the mechanical stimulation size.

[0034] In a second aspect, the embodiments of the present application provide an evaluation system for anal cleaning effect, comprising:

[0035] The anus anatomical structure, the multi-modal sensing assembly, the processor and the execution assembly are connected in sequence;

[0036] The multi-modal sensing assembly is configured to collect target parameters on the anus anatomical structure to obtain simulation monitoring data when the processor controls the execution assembly to simulate wiping of the test sample on the anus anatomical structure;

[0037] The processor is further configured to calculate a cleaning efficiency score according to the amount of residual dirt in the target monitoring data, and to call a target mechanical simulation model to calculate a skin irritation risk score according to the humidity penetration gradient value and the pressure value in the target monitoring data and the test sample, and to grade the skin irritation risk score according to a preset grade and combine the skin irritation risk score with the cleaning efficiency score to form a cleaning effect evaluation report of the test sample.

[0038] In some embodiments, the multi-modal sensing assembly is further configured to collect target parameters on the anus anatomical structure to obtain simulation monitoring data when the processor controls the execution assembly to simulate wiping of the test sample on the anus anatomical structure;

[0039] The processor is further configured to call an initial mechanical simulation model to perform finite element simulation based on initial model parameters to obtain comprehensive response data, and to compare the comprehensive response data with the simulation monitoring data to obtain calibration parameters to adjust skin material parameters in the initial model parameters to obtain the target mechanical simulation model.

[0040] The processor is further configured to call a reinforcement learning model to analyze the simulation monitoring data under different wiping strategies, to perform deep learning on obtained cleaning effect evaluation values and state spaces in the simulation monitoring data, to obtain reward values of different wiping strategies, and to update strategy parameters of a reward function of the reinforcement learning model to obtain the target wiping strategy.

[0041] In some embodiments, the anus anatomical structure comprises: simulation anus models connected to each other, a sphincter driving unit and a radial fold layer.

[0042] The sphincter driving unit comprises a pneumatic artificial muscle and a shape memory alloy wire; the pneumatic artificial muscle and the shape memory alloy wire are uniformly distributed around the anus of the simulated anus model to simulate the contraction of the circular sphincter; the radial wrinkle layer comprises a support base and a hydrogel simulated skin; the radial wrinkle layer is arranged on the support base, and the radial wrinkle layer is a ring-shaped radial groove structure for simulating the skin structure around the anus.

[0043] In some embodiments, the multi-modal sensing assembly comprises a pressure sensing array, an optical detector and a humidity gradient sensor;

[0044] The pressure sensing array is embedded in the hydrogel simulated skin in the form of a flexible piezoresistive film, for real-time acquisition of the pressure value of the radial wrinkle layer during simulated wiping of the radial wrinkle layer; the optical detector is arranged directly above the radial wrinkle layer, for real-time acquisition of the image of residual dirt on the radial wrinkle layer during simulated wiping of the radial wrinkle layer, to obtain the amount of residual dirt; the humidity gradient sensor is arranged in layers along the wrinkle depth direction of the hydrogel simulated skin, for real-time acquisition of the humidity penetration gradient value of liquid on the radial wrinkle layer during simulated wiping of the radial wrinkle layer.

[0045] Compared with the prior art, the application provides an anus cleaning effect evaluation method and system, which controls the execution assembly to wipe the anus anatomical structure according to the target wiping strategy of the to-be-tested sample, and controls the multi-modal sensing assembly to collect target parameters on the anus anatomical structure to obtain target monitoring data; according to the amount of residual dirt in the target monitoring data, a cleaning efficiency score is calculated, and a target mechanical simulation model is called, and according to the humidity penetration gradient value and the pressure value in the target monitoring data and the to-be-tested sample, a skin irritation risk score is calculated to form a cleaning effect evaluation report with the cleaning efficiency score, to realize simulation and prediction of the anus cleaning process, greatly improve the simulation test effect and accuracy, and give the cleaning effect evaluation report to explain the safety of the to-be-tested sample. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0047] Figure 1 A flow chart of the anus cleaning effect evaluation method provided by the application;

[0048] Figure 2 A flow chart for preliminarily obtaining the cleaning efficiency score and the skin irritation risk score in the evaluation method of the anal cleaning effect provided by the present application;

[0049] Figure 3 A flow chart for obtaining the cleaning efficiency score in detail in the evaluation method of the anal cleaning effect provided by the present application;

[0050] Figure 4 A flow chart for obtaining the skin irritation risk score in detail in the evaluation method of the anal cleaning effect provided by the present application;

[0051] Figure 5 A flow chart for the construction method of the mechanical simulation model provided by the present application;

[0052] Figure 6 A flow chart for obtaining the calibration parameters in the construction method of the mechanical simulation model provided by the present application;

[0053] Figure 7 A schematic diagram of the simulated wiping method for wiping the anus provided by the present application;

[0054] Figure 8 A structural schematic diagram of the evaluation system of the anal cleaning effect provided by the present application;

[0055] Figure 9 A distribution schematic diagram of the radial pleated layer and the sensor provided by the present application.

[0056] The figure reference: 10-anal anatomical structure; 11-simulated anus model; 12-sphincter driving unit; 13-radial pleated layer; 132-hydrogel simulated skin; 20-multimodal sensing assembly; 21-pressure sensing array; 23-humidity gradient sensor; 30-processor; 40-executing assembly. DETAILED DESCRIPTION

[0057] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments.

[0058] The components of the embodiments of the present application generally described and shown in the accompanying drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0059] Hereinafter, the terms "include", "have", and their conjugations, which are used in the various embodiments of the present application, merely indicate the presence of the features, numbers, steps, operations, elements, components, or combinations thereof, and do not preclude the presence or addition of one or more other features, numbers, steps, operations, elements, components, or combinations thereof.

[0060] Unless defined otherwise, all terms including technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which various embodiments of the present application belong. Such terms, as commonly used in the art, are thoroughly defined in commonly used dictionaries, and are not to be interpreted in an idealized or overly formal sense unless expressly so defined in the present application.

[0061] In the prior art, the disclosed anal model is a static structure that cannot simulate the contraction movement of the anal sphincter and the dynamic change of the anal folds; for example, the disclosed cleaning test method mainly relies on visual observation of liquid penetration, and lacks quantitative evaluation indicators.

[0062] The present application provides an evaluation method and system for anal cleaning effect. In the present application, when a test sample is used to simulate wiping of the anal anatomical structure according to a preset wiping strategy, target parameters are collected to obtain simulation monitoring data; after an initial mechanical simulation model is called to perform finite element simulation based on initial model parameters to obtain comprehensive response data, the initial model parameters are adjusted by comparing the simulation monitoring data with the comprehensive response data to obtain calibration parameters, so as to obtain a target mechanical simulation model; when a to-be-tested sample is used to simulate wiping of the anal anatomical structure according to a target wiping strategy, target parameters are collected to obtain target monitoring data, and the target monitoring data are analyzed to obtain a cleaning effect evaluation report of residual dirt, which realizes calibration of the parameters of the mechanical simulation model during simulation of the anal cleaning process, so as to obtain a more realistic mechanical simulation model, which is used for prediction of the anal cleaning process, so as to evaluate the anal cleaning effect, thereby effectively improving the simulation reality of the anal cleaning process and improving the accuracy of the evaluation of the cleaning effect.

[0063] It should be noted that the embodiments and features in the present application can be combined with each other in the case of no conflict. In addition, the specific steps included in the method described in the present application are not necessarily limited to the order described, but can be executed in any order or in parallel, unless expressly limited or contradictory in context.

[0064] The evaluation method design scheme of the anal cleaning effect will be described below through specific exemplary embodiments. It should be noted that the following embodiments are only used to explain and illustrate the technical solutions of the present application, and are not limited to specific embodiments:

[0065] Please refer to Figure 1 The evaluation method of the anal cleaning effect provided by the embodiments of the present application comprises steps S100-S300:

[0066] S100, control the execution assembly 40 to wipe the anal anatomical structure 10 according to the target wiping strategy, and control the multi-modal sensing assembly 20 to collect the target parameters on the anal anatomical structure 10, to obtain the target monitoring data.

[0067] In the present application, the test sample is taken as an example of wet toilet paper of different brands, which can also be dry toilet paper, etc.

[0068] Exemplarily, on the basis of obtaining the target mechanical simulation model and the target wiping strategy, the wet toilet paper (i.e. the test sample) of different brands is actually tested:

[0069] First, by the same reasoning, the execution assembly 40 is used to wipe the fluorescent marker dirt at the anal fold area on the surface of the radial fold layer 13 of the anal anatomical structure 10 according to the target wiping strategy, and the multi-modal sensing assembly 20 is used to correspondingly collect the pressure value, the amount of residual dirt and the humidity penetration gradient value and other target parameters on the anal anatomical structure 10 at this time, to obtain the target monitoring data.

[0070] S200, according to the amount of residual dirt in the target monitoring data, the cleaning efficiency score is calculated, and the target mechanical simulation model is called to calculate the skin irritation risk score according to the humidity penetration gradient value and the pressure value in the target monitoring data and the test sample.

[0071] Exemplarily, after obtaining the target monitoring data, the initial amount of dirt on the anal anatomical structure 10 and the amount of residual dirt in the target monitoring data are compared, and the cleaning efficiency score is calculated, for example, the depth removal rate, the dirt removal rate and the cleaning uniformity size and other data are calculated.

[0072] Then, the target mechanical simulation model is called to analyze the skin moisture balance, the chemical irritability, and the mechanical irritability according to the humidity penetration gradient value and the pressure value in the target monitoring data and the component information of the sample to be tested, so as to calculate the skin irritation risk score, so as to generate the cleaning effect evaluation report according to the cleaning efficiency score and the skin irritation risk score.

[0073] In one implementation method, referring to Figure 2 , step S200, the cleaning efficiency score is calculated according to the residual dirt amount in the target monitoring data, and the target mechanical simulation model is called to calculate the skin irritation risk score according to the humidity penetration gradient value and the pressure value in the target monitoring data and the sample to be tested, including:

[0074] S210, after the residual dirt amount on the radial fold layer 13 of the anal anatomical structure 10 in the target monitoring data is calculated, the depth removal rate, the dirt removal rate, and the cleaning uniformity are calculated, and then the cleaning efficiency score is calculated.

[0075] S220, the target mechanical simulation model is called to analyze the skin moisture balance and the chemical irritability according to the humidity penetration gradient value in the target monitoring data and the component information of the sample to be tested, and to evaluate the mechanical irritability according to the pressure value in the target monitoring data, so as to calculate the skin irritation risk score according to the skin moisture balance, the chemical irritability, and the mechanical irritability.

[0076] Demonstratively, after the target monitoring data is obtained, the cleaning efficiency score of the current sample to be tested is calculated:

[0077] The original dirt amount is compared with the residual dirt amount on the radial fold layer 13 of the anal anatomical structure 10 in the target monitoring data, for example, the original dirt amount and the residual amount of fluorescently labeled dirt on the radial fold layer 13 before and after the wiping action are compared to obtain the dirt removal rate, the dirt removal rate at the bottom of the radial fold layer 13 is compared, that is, the depth removal rate, and the size of the residual dirt distribution, so as to calculate the cleaning uniformity, and then the cleaning efficiency score is calculated.

[0078] Then, based on the humidity gradient sensor 23 data and the calibrated model parameters, the influence of the cleaning process on the skin moisture balance is analyzed:

[0079] The target mechanical simulation model is used to predict and evaluate the skin irritation caused by the component information of the sample to be tested (such as the content of surfactant), and the skin moisture balance is analyzed according to the actual humidity penetration gradient value reflected in the target monitoring data to analyze whether the skin is abnormally dry or has an irritation reaction, so as to reflect the chemical irritability.

[0080] Secondly, according to the pressure value of the radial fold layer 13 reflected in the target monitoring data, the local stress and friction damage risk assessment are analyzed to evaluate the mechanical stimulation size.

[0081] Finally, according to the skin moisture balance, the chemical stimulation size and the mechanical stimulation size, the skin stimulation risk score is calculated comprehensively, including risk level and specific risk factor analysis, etc., to serve as another important indicator of the cleaning effect evaluation report.

[0082] It can be understood that the present application can realize quantitative evaluation of the cleaning efficiency of the anal fold area, and significantly improve the accuracy and objectivity of the cleaning effect test results. Then, after calculating the depth removal rate, the dirt removal rate and the cleaning uniformity size according to the original dirt amount and the residual dirt amount after wiping, the cleaning efficiency score is calculated comprehensively to provide a basis for the cleaning report.

[0083] In one implementation method, please refer to Figure 3 According to the residual dirt amount on the radial fold layer 13 of the anal anatomical structure 10 in the target monitoring data, the depth removal rate, the dirt removal rate and the cleaning uniformity size are calculated, including:

[0084] S211, comparing the initial dirt amount at the fold bottom of the radial fold layer 13 with the residual dirt amount at the corresponding position in the target monitoring data to calculate the depth removal rate;

[0085] S212, comparing the initial total dirt amount on the radial fold layer 13 with the residual total dirt amount at the corresponding position to calculate the dirt removal rate;

[0086] S213, calculating the standard deviation size between the residual dirt amounts of different target areas in the radial fold layer 13 to obtain the cleaning uniformity size.

[0087] Exemplarily, the specific process of obtaining the cleaning efficiency score is as follows:

[0088] After collecting the fluorescence signal intensity at the fold bottom of the radial fold layer 13 after wiping as residual dirt, and collecting the residual total dirt on the radial fold layer 13 after wiping, the initial dirt amount at the fold bottom of the radial fold layer 13 is compared with the residual total dirt amount at the fold bottom in the target monitoring data to calculate the depth removal rate.

[0089] Then, in the same way, the initial total dirt amount of the radial fold layer 13 is compared with the residual total dirt amount on the radial fold layer 13 in the target monitoring data to calculate the dirt removal rate.

[0090] Wherein, if the residual dirt fluorescence intensity corresponds to a residual mass of less than 3 far below the set cleaning pass threshold (e.g. 10 ), by which the cleaning efficiency of the current tested wet toilet paper is determined to reach the pass standard.

[0091] Secondly, the standard deviation between the residual stain amount of different target areas (e.g. the top of the radial fold vs. the bottom of the radial fold) in the radial fold layer 13 is calculated, i.e. the standard deviation of the residual stain distribution in different target areas is calculated, to calculate the cleaning uniformity size. The smaller the standard deviation, the more uniform the cleaning, and at the same time, the “hot spot” area of the residual stain concentration can be identified as an important calculation area for calculating the cleaning uniformity size.

[0092] Finally, a comprehensive scoring model is used to calculate the final cleaning efficiency score (including overall score and sub-item score, with a full score of 100 and a passing line of 60) according to the depth removal rate, stain removal rate and cleaning uniformity size:

[0093] Cleaning efficiency score = w1*cleaning rate score + w2*uniformity score + w3*depth cleaning score;

[0094] where w1, w2 and w3 are weight coefficients of each index, reflecting the importance of different factors; the cleaning rate score is proportional to the stain removal rate; the uniformity score is inversely proportional to the standard deviation of the residual distribution; and the depth cleaning score reflects the cleaning effect at the deep part of the fold.

[0095] In one implementation method, please refer to Figure 4 , the target mechanical simulation model is called to analyze the skin moisture balance and the chemical irritability according to the humidity penetration gradient value in the target monitoring data and the composition information of the sample to be tested, and to evaluate the mechanical irritability according to the pressure value in the target monitoring data, including:

[0096] S221, according to the real-time change of the humidity penetration gradient value, the transdermal water loss rate and the change of skin water content are calculated to analyze the skin moisture balance;

[0097] S222, according to the humidity penetration gradient value, and the pH value and surfactant of the sample to be tested, the chemical irritability is predicted;

[0098] S223, according to the real-time change of the pressure value, the stress distribution of the surface and the inside of the radial fold layer 13 is calculated to evaluate the mechanical irritability.

[0099] Exemplarily, the specific evaluation process of the skin irritation risk score is as follows:

[0100] Firstly, according to the real-time change of the humidity penetration gradient value, the transdermal water loss rate is calculated by using the target mechanical simulation model, that is, the change of the water evaporation rate of the skin after cleaning is predicted, and the change of the skin water content is calculated, that is, the change of the water content of the skin at different depths before and after cleaning is evaluated, so as to analyze the skin moisture balance and analyze the influence of the cleaning process on the skin moisture balance. At the same time, when the transdermal water loss rate exceeds a certain multiple (for example, 2 times and above) of the normal value or the skin water content decreases by more than a threshold value in a short time, it is determined that there is a risk of skin dryness.

[0101] Then, the influence on the skin acid-base balance is evaluated according to the humidity penetration gradient value and the pH value of the sample to be tested by using the target mechanical simulation model, the influence on the skin barrier function is predicted according to the surfactant (such as SLS, sodium lauryl sulfate) contained in the sample to be tested, and the irritation reaction of the skin caused by certain ingredients (such as preservatives or fragrances) is analyzed, so as to predict the chemical irritation.

[0102] Secondly, according to the real-time change of the pressure value, the stress distribution on the surface and inside of the radial wrinkle layer 13 during wiping is calculated by using the target mechanical simulation model, so as to evaluate the mechanical irritation: if the local stress or friction force exceeds a safety threshold, it is determined that there is a risk of mechanical irritation. At the same time, the repeated wiping influence prediction is performed, that is, the cumulative damage risk caused by multiple repeated wiping is evaluated.

[0103] It can be understood that the skin irritation prediction for the pH value, surfactant concentration and other ingredient parameters of the wet toilet paper provides a scientific basis for the safety evaluation of the cleaning product.

[0104] S300, after the skin irritation risk score is graded according to the preset grade, the skin irritation risk score and the cleaning efficiency score constitute a cleaning effect evaluation report.

[0105] Demonstratively, after obtaining the skin irritation risk score and the cleaning efficiency score, the skin irritation risk score is first graded according to the risk level, and then written into the cleaning effect evaluation report together with the cleaning efficiency score, wherein the cleaning effect evaluation report further includes: sample information (such as the name, type and main ingredients of the sample to be tested), test conditions (such as test parameter settings, environmental conditions and simulated skin state), risk level determination, specific risk factor analysis and safe use suggestions (such as product use suggestions or improvement directions based on test results), etc., and the report can also include analysis and comparison, mainly the comparison with the results of other products or traditional test methods.

[0106] The risk level classification can include: low risk: all indicators are within the safe range, no obvious irritation risk; moderate risk: some indicators are close to the warning value, there is a potential irritation possibility; high risk: one or more indicators exceed the safety threshold, there is an obvious irritation risk.

[0107] Furthermore, the cleaning effect evaluation report can also be visually displayed:

[0108] For example, residual dirt distribution heat map: indicating the amount of residual dirt in different areas by color depth; cleaning efficiency curve: showing the relationship between the number of wipes and cleaning efficiency; skin stress distribution cloud map: showing the stress state of each part of the skin during wiping; moisture penetration gradient map: showing the penetration of liquid at different depths of the skin.

[0109] It can be understood that the present application establishes a comprehensive evaluation system including cleaning efficiency, uniformity, deep cleaning ability and skin safety, etc., which fully reflects the product performance. Moreover, subjective feelings are converted into objective data, such as converting fluorescence signal intensity into residual dirt amount and quantifying skin irritation risk into specific risk level, making the evaluation results more objective, accurate and comparable. Even, the evaluation standards and weights can be adjusted according to the needs of different user groups (such as sensitive skin groups and specific age groups), realizing personalized product evaluation and providing more targeted product recommendations for different groups of people.

[0110] In one implementation method, please refer to Figure 5 The control execution component 40 wipes the anal anatomical structure 10 according to the target wiping strategy, which includes S10-S30 before:

[0111] S10, the control execution component 40 simulates wiping the anal anatomical structure 10 according to the preset wiping strategy, and controls the multi-modal sensing component 20 to collect target parameters on the anal anatomical structure 10 to obtain simulation monitoring data.

[0112] In the present application, the test sample is a simple non-woven fabric / dust-free paper, which is used as a standard control group with the test sample, such as wet toilet paper (including: virgin pulp, non-woven fabric / dust-free paper, mild detergent (such as natural plant extract), antibacterial ingredients (such as tea tree essential oil) and risk additives (preservatives (such as methylisothiazolinone), fragrances or bactericides (such as benzalkonium chloride)) etc.), and a certain amount of fluorescently labeled dirt (for example, a mixture of appropriate amount of fluorescent dye and simulated fecal material) is uniformly applied on the anal fold area on the surface of the radial fold layer 13 of the anal anatomical structure 10 to simulate dirt, and the fluorescence signal intensity is used as the amount of dirt. The target parameters include: pressure value, residual dirt amount and moisture penetration gradient value.

[0113] Exemplarily, the application needs to optimize the mechanical simulation model before evaluating the anal cleaning effect:

[0114] The control execution component 40 uses the preset wiping strategy to simulate wiping of the fluorescently labeled dirt on the anal anatomical structure 10 with non-woven fabric / dusting paper (i.e., the test sample), to simulate wiping of the test sample on the dirt on the anal anatomical structure 10.

[0115] Then, the various sensors in the multi-modal sensing component 20 collect target parameters such as pressure values, residual dirt amounts, and humidity penetration gradient values on the anal anatomical structure 10, to obtain simulated monitoring data.

[0116] It can be understood that, in the application, the anal anatomical structure 10 is simulated and wiped, and relevant target parameters are collected, to adjust the anal anatomical structure 10, to simulate a real human anus, so as to obtain more accurate test data.

[0117] Further, in another embodiment of the application, the simulated monitoring data is also preprocessed and formatted:

[0118] First, the simulated monitoring data is cleaned to filter out noise and outliers, such as sudden peaks that may exist in the pressure sensing array or fluctuation interference of the humidity sensor; then, a spatio-temporal alignment operation is performed: the data of different sensors at the same time point and the same spatial position are aligned to establish a spatio-temporally consistent data set; then, data format conversion is performed, i.e., the simulated monitoring data is converted into an input format acceptable to the mechanical simulation model (such as a finite element model), such as converting the pressure distribution data into nodal forces or boundary conditions.

[0119] S20, after calling the initial mechanical simulation model to perform finite element simulation based on the initial model parameters to obtain comprehensive response data, comparing the simulated monitoring data, to obtain the calibration parameters.

[0120] Among them, the mechanical simulation model includes: the initial mechanical simulation model and the target mechanical simulation model, the mechanical simulation model is the finite element model, which is used to simulate the mechanical response and transmission process of the anal wrinkle skin under the conditions of stress and wetting.

[0121] Exemplarily, after obtaining the simulated monitoring data, the simulated monitoring data is input into the initial mechanical simulation model for parameter inversion:

[0122] First, the initial mechanical simulation model is constructed, i.e., a digital twin containing initial model parameters is constructed. Among them, the initial model parameters include: geometric structure, material properties, boundary conditions, load conditions, etc. Among them, the material properties include: elastic modulus (such as about 20 kPa), initial liquid permeability, initial friction coefficient, and other related parameters (such as Poisson's ratio, density, etc.).

[0123] Then, the finite element solver is called to perform multiple forward finite element simulations based on the initial model parameters to simulate the response behavior of the anal anatomical structure 10 under given inputs, collect and output the comprehensive response data such as stress-strain distribution, temperature distribution, pressure value, residual contamination amount and humidity permeation gradient value, etc., until the objective function converges (e.g., the number of iterations reaches the maximum number or the difference < 0.01).

[0124] Finally, the simulation results are compared with the simulated monitoring data to obtain the calibration parameters for the correction, error evaluation and iterative optimization of the digital twin, etc.

[0125] It can be understood that in the present application, the difference between the measured data and the prediction results of the model is compared, and the model parameters are adjusted by reverse deduction until the model output is maximally matched with the measured data. Then, the calibration parameters are obtained by comparing the latest predicted comprehensive response data with the simulated monitoring data, so as to adjust the skin material parameters in the initial model parameters of the initial mechanical simulation model, so as to simulate the real wiping action, and finally improve the accuracy of predicting the cleaning efficiency of the wet toilet paper.

[0126] S30, adjusting the skin material parameters in the initial model parameters according to the calibration parameters to obtain a target mechanical simulation model.

[0127] The skin material parameters include the equivalent liquid permeability, elastic modulus and friction coefficient of the skin layer.

[0128] Exemplarily, after the calibration parameters are calculated, the corresponding parameters in the skin material parameters are adjusted according to the calibration coefficients of the friction, elastic modulus and equivalent liquid permeability, respectively, to obtain a target mechanical simulation model which can accurately reflect the parameters of the real skin tissue characteristics.

[0129] In one implementation method, please refer to Figure 6 , after step S20, the initial mechanical simulation model is called to perform finite element simulation based on the initial model parameters to obtain comprehensive response data, and the calibration parameters are obtained by comparing the simulated monitoring data, including:

[0130] S21, the initial mechanical simulation model is called to perform multiple forward finite element simulations based on the initial model parameters until the number of iterations reaches the preset maximum number of iterations to obtain the predicted friction and the predicted equivalent liquid permeability;

[0131] S22, after the actual friction is analyzed and obtained according to the pressure distribution change in the simulated monitoring data and the motion parameters of the execution assembly 40, the friction and the elastic modulus are compared with the predicted friction to obtain the reference coefficients of the friction and the elastic modulus;

[0132] S23, analyze the penetration speed, distribution and moisture change data of the liquid in the different depths of the skin in the simulation monitoring data, obtain the actual equivalent liquid permeability of the skin layer, and compare the actual equivalent liquid permeability with the predicted equivalent liquid permeability to obtain a calibration coefficient of the equivalent liquid permeability.

[0133] The motion parameters include speed, acceleration, joint torque, position, and motion direction.

[0134] Exemplarily, the specific process of obtaining the calibration parameter by using parameter inversion is as follows:

[0135] First, after obtaining the simulation monitoring data, a three-dimensional geometric model is constructed based on the actual shape of the radial fold layer 13, including the fold depth and fold distribution and other characteristics. The geometric model is discretized into a finite element grid, a coarse grid is used for non-critical areas to save computing resources, and a finer grid is used for the fold area to improve the calculation accuracy, initial material parameters (including geometric structure, material properties, boundary conditions, load conditions, etc.) are set for different parts in the geometric model (such as the support base and the hydrogel simulation skin 132), and finally a digital twin model (i.e. a mechanical simulation model, i.e. a finite element model) containing initial model parameters is obtained. Among them, the boundary condition setting: set the boundary constraints of the model according to the experimental conditions, such as fixed support points, contact surfaces, etc. Load condition definition: the pressure distribution measured by the pressure sensor array 21 is input as the external load of the model.

[0136] At the same time, a target function reflecting the difference between the model prediction and the measured data is constructed:

[0137] (1);

[0138] In formula (1), F(p) is the target function to be minimized; p is the parameter vector to be inverted (including liquid permeability, friction coefficient, etc.); Mi(p) is the predicted value of the model using the parameter p at measuring point i; Di is the measured data at measuring point i; wi is the weight coefficient of each measuring point.

[0139] Then, the sensitivity of the target function to each parameter is calculated to determine the importance ranking of the parameters, and the parameters with high sensitivity have a greater impact on the model prediction results and are preferentially inverted.

[0140] Then, the mechanical simulation model performs multiple forward finite element simulations based on the initial model parameters, iteratively solving for the predicted friction force and the predicted equivalent liquid permeability, i.e., running the mechanical simulation model with the initial parameters p0, obtaining the initial prediction result M(p0), calculating the difference between the prediction result and the measured data, evaluating the objective function value F(p0), updating the parameter value according to the optimization algorithm, obtaining the new parameter p1, running the model again using the new parameter p1, obtaining the updated prediction result M(p1), calculating the new objective function value F(p1), and comparing it with F(p0) of the previous step: if |F(p1)-F(p0)|<ɛ (a preset convergence threshold, such as 0.01) or the number of iterations reaches a preset maximum number of iterations (e.g., the number of iterations does not exceed 10), stop the iteration; otherwise, continue the iteration. When the iteration is completed, the finite element model is run with the optimal parameters, and after the pressure distribution is integrated, the theoretical friction force value, i.e., the predicted friction force, can be calculated by combining the inverted friction coefficient. At the same time, the finite element model also simulates the penetration process of the liquid in each layer of the skin, outputs the concentration distribution, moisture content and penetration depth at any time point, and analyzes to obtain the predicted equivalent liquid permeability.

[0141] The iteration process and iteration record are shown in Table 1 as follows:

[0142]

[0143] Table 1

[0144] Wherein, at convergence: iterate until the difference ɛ<=0.01, use gradient descent to optimize the objective function , added to the software interface description (such as Python script calling finite element solver).

[0145] Then, according to the pressure distribution change monitored by the pressure sensing array 21 in the multi-modal sensing assembly 20, and the motion parameters (including speed, acceleration, joint torque, position and motion direction, etc.) of the execution assembly 40, and combining with the dynamics model to calculate the external force on the end effector, the actual contact force and the actual friction force are inversely calculated. Then, the actual friction force and the predicted friction force are compared, and the benchmark coefficient of the friction on the skin surface is inverted by minimizing the difference between them.

[0146] And, the penetration speed of the liquid at different depths of the skin, the distribution of the liquid and the moisture change data monitored by the humidity gradient sensor 23 in the multi-modal sensing assembly 20 are input into the initial mechanical simulation model, the time and spatial distribution characteristics of the liquid penetrating from the surface to the deep layer are analyzed, and the actual equivalent liquid permeability of the skin layer is obtained. Then, the ratio of the predicted equivalent liquid permeability to the actual equivalent liquid permeability is obtained to obtain the calibration coefficient of the equivalent liquid permeability.

[0147] In an implementation method, please refer to Figure 7 , the control execution component 40 wipes the anus anatomical structure 10 according to the target wiping strategy, which further includes:

[0148] S40, call the target mechanical simulation model to analyze the simulated monitoring data collected under different wiping strategies to obtain the cleaning effect evaluation value.

[0149] Among them, the simulated monitoring data and the target mechanical simulation model are obtained according to the above method of constructing the mechanical simulation model; the wiping strategy includes: wiping path, wiping force, wiping speed, wiping angle and wiping times, etc.

[0150] Exemplarily, the simulated monitoring data collected under different wiping strategies (wiping path, wiping force, wiping speed, wiping angle and wiping times, etc.) are obtained by the same method in the method of constructing the mechanical simulation model, and the target mechanical simulation model is used to analyze different simulated monitoring data respectively:

[0151] According to the amount of residual dirt on the radial fold layer 13 of the anus anatomical structure 10 reflected on the simulated monitoring data, the cleaning efficiency score is solved, and according to the humidity penetration gradient value reflected by the target monitoring data, the skin irritation risk score is calculated, and according to the two scores, the cleaning effect evaluation value of different wiping strategies is calculated.

[0152] It can be understood that in this application, by using the target mechanical simulation model to analyze different simulated monitoring data to obtain the cleaning effect evaluation value, the reinforcement learning model is used as an important basis for deep learning to optimize the wiping strategy, so as to effectively improve the wiping effect of the wiping strategy, which not only helps to improve the wiping cleanliness but also reduces the wiping damage, and thus effectively improves the wiping comfort.

[0153] S50, call the reinforcement learning model to perform deep learning on the state space in the simulated monitoring data and the cleaning effect evaluation value to obtain the reward value of different wiping strategies.

[0154] Among them, the state space is monitored by the corresponding sensor in the multi-modal sensing component 20; the state space includes: the surface dirt residual distribution of the radial fold layer 13 in the anus anatomical structure 10 (monitored by the optical detector), the fold deformation state (monitored by the pressure sensing array 21) and the internal humidity gradient distribution (monitored by the humidity gradient sensor 23), etc.

[0155] wherein, the wiping path: the trajectory of the end of the mechanical arm on the skin surface, which can be a straight line, a curve or a composite path; the wiping force: the normal pressure applied by the mechanical arm on the skin surface, which can be in the range of 0.5-3.0 kPa; the wiping speed: the moving speed of the end of the mechanical arm, which can be about 50 mm / s; the wiping angle: the contact angle of the cleaning material with the skin surface; the wiping times: the number of repeated wiping.

[0156] Exemplarily, after the cleaning effect evaluation value is obtained, the state space in the cleaning effect evaluation value and the simulation monitoring data is subjected to deep learning by using the reinforcement learning model in the processor 30:

[0157] A reasonable reward function is designed first, including: cleaning efficiency reward, skin safety reward, resource efficiency reward and comprehensive score reward, etc.; wherein, the cleaning efficiency reward: according to the amount of residual dirt measured by the optical detection module, the higher the dirt removal rate, the greater the reward. For example, when the fluorescence intensity of the residual dirt is lower than 3 (the qualified standard), a higher reward is given. The skin safety reward: according to the stress state of the skin predicted by the pressure sensing array 21 and the finite element model, when the pressure distribution is uniform and lower than the safety threshold (the threshold at which skin tissue damage is likely to occur), a positive reward is given; otherwise, when the pressure is too high and may cause skin damage, a negative reward is given. The resource efficiency reward: related to the number of wipes required to achieve the same cleaning effect, the fewer the number of wipes, the higher the reward, encouraging the algorithm to find the most efficient cleaning strategy. The comprehensive score reward: the “cleaning efficiency score” is calculated by the comprehensive score test formula, the higher the score, the greater the reward.

[0158] Then, the state space and other data measured under the current wiping strategy are input into the reinforcement learning algorithm (such as deep Q learning network or policy gradient method, etc.) of the reinforcement learning model for deep learning, to calculate the reward value of the current cleaning state-wiping action combination.

[0159] S60, update the policy parameters of the reward function of the reinforcement learning model according to the reward value, to obtain the target wiping strategy.

[0160] Exemplarily, after the reward values of different wiping strategies are obtained, the Q values or policy parameters of the reinforcement learning algorithm are updated, and the wiping path and the applied force of the mechanical arm are iteratively optimized according to the cleaning effect evaluation value under the current test strategy, until the optimal wiping strategy (i.e. the target wiping strategy) is found, i.e. the optimal wiping path, wiping speed and wiping pressure combination method that minimizes the residual dirt is found.

[0161] It can be understood that the reward values of different wiping strategies are obtained by deep learning of the cleaning effect evaluation value and state space and other data by the reinforcement learning model to update the strategy parameters of the reward function of the reinforcement learning model, so as to obtain the target wiping strategy, and the automatic optimization of the wiping strategy is realized. On the basis of reducing the wiping times and the amount of cleaning agent, the residual dirt is minimized, the risk of skin irritation is minimized, and the resource utilization is optimized.

[0162] Further, in another embodiment provided in the present application, the influence of a single factor (such as different wiping pressures or different skin states) on the cleaning effect in different wiping strategies can also be simulated respectively. For example, the best wiping pressure of 1.5 kPa is found through comparative experiments.

[0163] Please refer to Figure 8 The anus cleaning effect evaluation system provided in the embodiments of the present application comprises:

[0164] The anus anatomical structure 10, the multi-modal sensing assembly 20, the processor 30 and the execution assembly 40 are sequentially connected.

[0165] The multi-modal sensing assembly 20 is configured to collect target parameters on the anus anatomical structure 10 to obtain simulation monitoring data when the processor 30 controls the execution assembly 40 to simulate wiping of the test sample on the anus anatomical structure 10.

[0166] The processor 30 is configured to compare the simulation monitoring data with comprehensive response data obtained by executing finite element simulation based on initial model parameters by calling an initial mechanical simulation model, and obtain calibration parameters to adjust skin material parameters in the initial model parameters.

[0167] The processor 30 is further configured to perform deep learning on the simulation monitoring data by calling a reinforcement learning model to update strategy parameters and obtain a target wiping strategy.

[0168] The processor 30 is further configured to calculate a cleaning efficiency score according to the amount of residual dirt in the target monitoring data, and call a target mechanical simulation model to calculate a skin irritation risk score according to the humidity penetration gradient value and the pressure value in the target monitoring data and the test sample, and then grade the skin irritation risk score according to a preset grade to form a cleaning effect evaluation report of the test sample together with the cleaning efficiency score.

[0169] The anus anatomical structure 10 and the multi-modal sensing assembly 20 are directly connected by a physical method, and the processor 30 is connected with the multi-modal sensing assembly 20 and the execution assembly 40 by signals; the execution assembly 40 is a mechanical arm, which can be a six-degree-of-freedom industrial robot, and the end thereof is fixed with a clamp for clamping cleaning materials (such as dry toilet paper or wet toilet paper).

[0170] Exemplarily, the system of the present embodiment corresponds to the evaluation method of the anal cleaning effect of the above-mentioned embodiment, and the optional items in the above-mentioned embodiment are also applicable to the present embodiment, and thus are not repeated here.

[0171] It can be understood that, in the test process, the execution assembly 40 simulates the wiping action of the dynamic anatomical structure according to the preset program. Synchronously, the multi-modal data (including the stress distribution, residual dirt image, humidity gradient, etc.) collected by the multi-modal sensing assembly 20 is input into the finite element model (i.e., the mechanical simulation model) in the processor 30 in real time for processing. The finite element model continuously calibrates the skin material parameters (such as liquid permeability, friction coefficient, etc.) of itself according to the measured data to ensure that the simulation results are consistent with the physical test. At the same time, the reinforcement learning model in the processor 30 iteratively optimizes the wiping path and the applied force of the mechanical arm based on the data measured by the finite element model and the sensor, that is, by analyzing the current cleaning effect through the reinforcement learning algorithm, comparing it with the target cleaning degree, and continuously adjusting the mechanical arm action to find the optimal wiping strategy, i.e., the target wiping strategy.

[0172] Moreover, the present application reflects the closed-loop process of feedback control of the mechanical arm through the digital twin model processing of the sensor data, embodying the real-time fusion of the physical system and the digital system, i.e., the physical experiment provides real data support for the digital model, and then the digital model in turn optimizes the execution of the physical system, thereby realizing the autonomous optimization and verification of the cleaning process.

[0173] Finally, compared with the traditional artificial-dependent cleaning effect test method, the evaluation system of the anal cleaning effect in the present application can reduce the test cost by about 60%, shorten the test period by about 75%, greatly improve the test efficiency and reduce the resource consumption; finally, the simulation platform provided by the present application has good universality and repeatability, and can simulate different types of cleaning products (such as dry toilet paper, wet toilet paper, wet wipes, etc.), different skin conditions and various wiping methods by adjusting parameters, ensuring that the test results have good repeatability and comparability, thereby comprehensively improving the reliability and efficiency of the development and test of anal cleaning products.

[0174] In one implementation method, the anal anatomical structure 10 includes: a simulated anal model 11, a sphincter driving unit 12 and a radial pleated layer 13 connected to each other in pairs.

[0175] The sphincter driving unit 12 includes: pneumatic artificial muscles and shape memory alloy wires; the pneumatic artificial muscles and shape memory alloy wires are uniformly distributed around the simulated anal model 11 to simulate the contraction of the circular sphincter; the radial wrinkle layer 13 includes: a support base and a hydrogel simulated skin 132; the radial wrinkle layer 13 is arranged on the support base, and the radial wrinkle layer 13 is a ring-shaped radial groove structure for simulating the skin structure around the anus.

[0176] Exemplarily, a plurality of pneumatic artificial muscles and shape memory alloy wires are uniformly distributed along the circumference of the simulated anal model 11 in the anal anatomical structure 10; the sphincter driving unit 12 adopts a driving mode combining pneumatic artificial muscles and shape memory alloy wires, the pneumatic artificial muscles contract when inflated to generate the required mechanical tension, and the shape memory alloy wires contract when heated by electricity to assist in providing rapid and subtle contraction force adjustment. By adjusting the air pressure and current, the two actuating elements cooperatively output a total contraction force of 0.5-5 N and can achieve a periodic contraction frequency in the range of 0.1-2 Hz, realizing the opening and closing movement of the sphincter, thereby realistically simulating the contraction-relaxation movement of the real anal sphincter.

[0177] The radial wrinkle layer 13 simulates the plicated skin structure around the anus, which can be formed by 3D printing with light-cured resin into a ring-shaped radial groove structure, and the surface is covered with a simulated hydrogel “skin” (i.e., hydrogel simulated skin 132) with an elastic modulus of about 20±5 kPa, close to the softness of human skin, maintaining a certain water content to simulate the wet characteristics of the skin, and being provided with adjustable wrinkle depth (e.g., 0.2-1.5 mm) to restore the structural characteristics of the human anal plication, thereby simulating the plication deformation of the real anus.

[0178] In one implementation method, the multi-modal sensing assembly 20 includes: a pressure sensing array 21, an optical detector, and a humidity gradient sensor 23.

[0179] The pressure sensing array 21 is embedded in the form of a flexible piezoresistive film inside the hydrogel simulated skin 132, for real-time acquisition of the pressure value of the radial wrinkle layer 13 when simulating wiping of the radial wrinkle layer 13; the optical detector is arranged directly above the radial wrinkle layer 13, for real-time acquisition of the image of the residual dirt on the radial wrinkle layer 13 when simulating wiping of the radial wrinkle layer 13, to obtain the amount of residual dirt; the humidity gradient sensor 23 is arranged in layers along the wrinkle depth direction of the hydrogel simulated skin 132, for real-time acquisition of the humidity penetration gradient value of the liquid on the radial wrinkle layer 13 when simulating wiping of the radial wrinkle layer 13.

[0180] Exemplarily, in combination with Figure 9The radial fold layer 13 is shown with a distribution of sensors. A pressure sensor array 21 is arranged in a grid pattern under the surface of the radial fold layer 13. The pressure sensor array 21 is used to detect the contact pressure at different positions in real time when the mechanical arm wipes the radial fold layer 13 with wet toilet paper. The resolution of the pressure sensor array 21 can reach about 1 kPa, and the response time is less than 10 ms, so that the instantaneous pressure change can be captured.

[0181] An optical detector is installed above the radial fold layer 13 and aligned with the skin surface. It works by irradiating near-infrared light (wavelength about 850 nm) on the folded surface and collecting the fluorescence signal image by the CMOS image sensor above, to detect the residual distribution of pre-applied fluorescently labeled dirt, and to collect the image of residual dirt on the radial fold layer 13 in real time to obtain the amount of residual dirt.

[0182] The humidity gradient sensor 23 is arranged in layers along the depth of the fold. It is composed of impedance measurement units made up of several groups of interdigital electrodes, which are arranged at different depths with an accuracy of about ±3% relative humidity. For example, a humidity sensor is installed every 1 mm along the depth of the fold to measure the liquid content or humidity gradient at different depths inside the fold, detect the penetration and diffusion of liquid at different depths under the skin, and draw the gradient distribution of humidity along the depth. This distributed layout can obtain the drying degree variation from top to bottom of the cleaned area after wiping, and determine whether the cleaning liquid has penetrated into the deep part of the fold, providing a quantitative basis for evaluating liquid penetration and residue.

[0183] Then, by inputting the pressure distribution provided by the pressure sensor array 21, the liquid penetration data provided by the humidity sensor, and other data into the finite element model, the skin material parameters consistent with the actual measurement are obtained by inversion. After the model parameter calibration is completed, the digital twin model (i.e. the mechanical simulation model) can accurately predict the interaction between the skin and the cleaning agent under different cleaning conditions. The mechanical arm acts as an execution mechanism (i.e. the execution assembly 40) and forms a physical-digital closed loop with the above-mentioned digital twin model. During the test process, the mechanical arm combines the feedback results of the digital twin model to continuously optimize the wiping strategy. For example, through autonomous training of the reinforcement learning model, the system can find the optimal wiping path and pressure combination that minimizes residual dirt, i.e. the target wiping strategy. Then, the mechanical arm executes the wiping action according to the target test strategy, while the sensors collect new data and feed back to the digital twin model, and so on, to realize intelligent control and dynamic verification of the cleaning process (information flow closed loop relationship). This physical and digital combined verification mechanism ensures that the simulation system can adjust itself according to different cleaning schemes, and accurately predict the cleaning effect under various working conditions.

[0184] The following examples further illustrate the application and effectiveness of the method for evaluating the cleaning effect of the anus:

[0185] Example 1, testing the cleaning efficiency of a certain brand of wet toilet paper:

[0186] This example uses the above-mentioned anal cleaning effect evaluation system to simulate and test the cleaning effect of a certain brand of wet toilet paper on the market.

[0187] First, evenly apply a layer of simulated dirt (containing fluorescent markers) to the anal fold area on the surface of the radial fold layer 13. After the simulated dirt dries and adheres, set the fold depth of the radial fold layer 13 to 0.8 mm, and apply a circumferential contraction force of about 2 N through the sphincter driving unit 12 to simulate the skin fold state when the anal sphincter is moderately contracted. Subsequently, the end of the mechanical arm holds the wet toilet paper and cleans the fold area according to the target wiping strategy: the mechanical arm performs a back-and-forth wiping motion on the skin surface at a linear speed of about 50 mm / s and applies a normal pressure of about 1.5 kPa, repeating the motion 3 times (equivalent to 6 single-pass wipes).

[0188] Then, during the entire wiping process, the multi-modal sensing assembly 20 synchronously collects data: the pressure sensing array 21 records the pressure distribution on each part of the skin during each wipe, and the results show that the pressure is basically uniform and remains in the range of 1.4-1.6 kPa, proving that the mechanical arm is effectively controlled according to the set pressure; the humidity gradient sensor 23 monitors that due to the moisture contained in the wet toilet paper itself, the skin surface layer is slightly more humid after wiping, but the humidity in the deep folds (>1 mm) changes little, indicating that the cleaning liquid mainly stays on the superficial layer; the optical detector captures the fluorescent image of the skin surface after each wipe. After 3 back-and-forth wipes, the optical detection results show that only a small amount of fluorescent dirt remains on the skin surface. Quantitative analysis shows that the fluorescent intensity of the remaining dirt corresponds to a residual material mass of less than 3 , far below the set cleaning pass threshold (e.g., 10 ). Therefore, the system determines that the cleaning efficiency of the wet toilet paper meets the qualified standard.

[0189] Further, according to the reduction ratio of the remaining dirt, the system calculates that the cleaning efficiency score of the product is about 95 points (out of 100 points, with higher scores indicating better cleaning effect). At the same time, since the wet toilet paper formula does not contain irritating ingredients (its pH value is close to neutral and does not contain high-concentration surfactants), neither the humidity sensing data nor the digital twin model shows signs of abnormal drying or irritation of the skin, so the system's skin irritation risk assessment report shows a low risk level. As can be seen, the anal cleaning effect evaluation system can objectively evaluate the cleaning performance of wet toilet paper under laboratory conditions, and the results show that the tested wet toilet paper can effectively clean the anal fold area under the set conditions and does not cause significant irritation to the skin.

[0190] Example 2, Evaluation of the skin irritation of surfactants in cleaning agents:

[0191] This example uses the digital twin simulation capabilities of the anal cleaning effect evaluation system to predict and evaluate the skin irritation that may be caused by the content of surfactants in a certain cleaning formula.

[0192] First, sodium dodecyl sulfate (SLS) is selected as a representative surfactant, and a concentration of 0.5% SLS is set in the cleaning liquid in the mechanical simulation model (such as a finite element model). Based on literature and experimental data, the skin permeation parameters corresponding to 0.5% SLS solution (such as the wettability change caused by the stratum corneum under this concentration, the increase in lipid extraction rate, etc.) are input into the finite element model.

[0193] Then, the SLS-containing cleaning liquid is simulated to be coated and left on the surface of the radial wrinkle layer 13 by a mechanical arm, and the moisture change data at different depths of the skin is obtained by the humidity gradient sensor 23; at the same time, the finite element model calibrates the skin permeation parameters according to the initial feedback of the sensor. Then, the calibrated finite element model is used to simulate the skin moisture loss over a period of time (for example, 30 minutes). The simulation results show that the cleaning liquid containing 0.5% SLS significantly increases the skin surface moisture loss rate, which is much higher than the safety threshold. For example, the model calculates that the trans-epidermal water loss rate (TEWL) continues to exceed twice the normal value, and the skin moisture content decreases by more than the threshold level in a short time, indicating that the skin barrier function may be damaged. According to the pre-set irritation judgment standard (for example, if the TEWL rises by more than a certain percentage, it is judged to have a risk of irritation), the system automatically triggers a warning prompt for skin irritation, considering that the formula has a high risk of skin irritation.

[0194] As can be seen, through the anal cleaning effect evaluation system, without testing on real human skin, only through simulation and a small amount of sensor data calibration, the skin irritation that may be caused by chemical components in the cleaning formula can be predicted, providing a basis for guiding the safety improvement of cleaning products.

[0195] Example 3, Comparison test of cleaning effects of different cleaning products:

[0196] This example uses the anal cleaning effect evaluation system to compare and test the differences in cleaning effects of two different cleaning methods under the same conditions: one is ordinary dry toilet paper wiping, and the other is pre-wetted wet toilet paper wiping.

[0197] First, two skin models of the radial wrinkle layer 13 with the same initial conditions (the wrinkle depth is set to 1.0 mm, and the water gel skin layer state is consistent) are prepared before the experiment. The same formula and amount of simulated dirt are uniformly coated on the surface of each skin, and allowed to adhere and dry.

[0198] Then, the same mechanical arm was used to clean both skin models with dry toilet paper and wet toilet paper respectively, with the same wiping parameters: same wiping path and area, normal pressure of 1.5 kPa, wiping speed of 50 mm / s, and 2 times of back-and-forth wiping. During the test, the two sets of sensors recorded data, and the amount of residual dirt after wiping was detected by optical detection.

[0199] Secondly, there was a significant difference in the test results of the two cleaning products: after wiping with dry toilet paper, more fluorescent dirt remained on the skin surface, and the residual mass corresponding to the fluorescent intensity value was about 20 While the residual dirt after wiping with wet toilet paper was significantly reduced, with a residual mass of about 4 , which was only one-fifth of the former. This shows that wet toilet paper can more effectively dissolve and remove dirt due to its own moisture and cleaning ingredients.

[0200] However, in the data of the pressure sensor array 21, the pressure distribution and size were similar in both cases, which means that the difference in cleaning effect mainly comes from the different properties of the cleaning materials themselves. The data of the humidity gradient sensor 23 shows that the skin surface humidity increased slightly after wiping with wet toilet paper, and the cleaning liquid was detected to have penetrated slightly into the folds (about 5-8%RH), while wiping with toilet paper did not change the internal humidity of the skin (as no additional liquid was introduced). This further illustrates that wet toilet paper brings cleaning liquid into the folds, which helps to clean the dirt in the deep part. Combining the optical detection results, the system calculates that the cleaning efficiency score of the wet toilet paper scheme is about 90, while that of the dry toilet paper scheme is only about 70, and wet toilet paper is significantly better than dry toilet paper.

[0201] And in terms of skin irritation risk, since dry toilet paper does not contain chemical reagents, its chemical irritation risk is zero, but the mechanical arm recorded that the skin friction was slightly higher when wiping dry, and long-term repeated dry wiping could cause slight skin abrasions; in contrast, wet toilet paper reduces the friction coefficient and brings moisturizing effect, so neither of them caused obvious skin irritation alarms in this short-term test.

[0202] Therefore, through this comparative test, the differences in performance of different cleaning products can be directly observed, and the system can be used to quantitatively compare the pros and cons of various cleaning schemes, providing data support for consumer product selection and manufacturer formula improvement.

[0203] Example 4, Effect of Different Wiping Pressures on Cleaning Effect Test:

[0204] This example uses the system to study the effect of wiping pressure, a key parameter, on cleaning effect and skin safety.

[0205] First, the experiment uses the same brand of wet toilet paper as the cleaning material, and compares the wiping tests under different applied pressures on the same skin model. Prepare a skin model with a wrinkle depth of 1.0 mm, and coat the surface with a fixed amount of fluorescent agent containing simulated dirt. The mechanical arm presets three different wiping schemes: low pressure scheme (normal pressure about 0.5 kPa), medium pressure scheme (1.5 kPa) and high pressure scheme (3.0 kPa). Except for the pressure size, other parameters remain the same (such as the same wiping path, speed about 50 mm / s, single wiping once).

[0206] Then, the three schemes are executed in turn, and the optical and humidity sensor data are restored to the initial state after each execution, and then the same dirt is re-coated for the next test to ensure that the test conditions are repeatable. The experimental results show that: at a low pressure of 0.5 kPa, the wet toilet paper does not contact the skin closely, and a large amount of dirt remains after wiping, with an optical detection residual amount of more than 50% of the initial dirt amount; when the pressure is increased to 1.5 kPa, the dirt is basically cleaned up, the residual amount is reduced to less than 5% of the initial (most areas have no fluorescent signal, only a few areas detect a small amount of residual), and the cleaning effect is significantly improved; when the pressure is further increased to 3.0 kPa, the residual dirt is further reduced to about 2%, which is not much higher than that at 1.5 kPa, indicating that the pressure is too high and has entered the saturation range of the cleaning effect.

[0207] At the same time, the pressure sensor array 21 records show that the peak pressure of each measurement point of the skin under the 3.0 kPa scheme is twice that of the 1.5 kPa, and the local pressure is even close to the threshold of skin tissue damage. The digital twin model further predicts that under the 3.0 kPa pressure wiping, if repeated multiple times, it may cause a risk of slight damage to the skin surface layer; while 1.5 kPa is within the safe range.

[0208] And, considering the cleaning effect and safety, the system analysis considers that the medium pressure (about 1.5 kPa) is the best wiping pressure: it can remove most of the dirt in one wiping, and will not cause obvious damage risk to the skin.

[0209] Then, this embodiment proves that the system can simulate and analyze the impact of wiping parameters on cleaning effect and skin health, providing a basis for formulating optimized cleaning operation parameters. For example, manufacturers can adjust the recommended wiping force in the product usage instructions to balance the cleaning force and the friendliness to the skin.

[0210] In summary, all the above examples fully demonstrate the feasibility and effectiveness of the evaluation system and method of anal cleaning effect. Through the present application, researchers can simulate the real anal cleaning process in a laboratory environment, obtain quantitative cleaning efficiency indicators and skin reaction data, thereby accelerating the development and testing of cleaning products and improving the reliability of test results. Those skilled in the art can make various equivalent changes and adjustments according to the idea of the present application, such as changing the sensor type, replacing the driving mechanism or adjusting the algorithm parameters, etc., which all fall within the protection scope of the present application.

[0211] The present application also provides a computer device, which exemplarily comprises a processor and a memory, wherein the memory stores a computer program, and the processor executes the functions of each module of the above-mentioned PWS-based early warning method or the above-mentioned PWS-based early warning system by running the computer program.

[0212] The processor can be an integrated circuit chip with a processing capability of signals. The processor can be a general processor, including a central processing unit (CPU), a graphics processing unit (GPU), and a network processor (NP), a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, a discrete gate or transistor logic device, a discrete hardware component, or at least one of the above. The general processor can be a microprocessor or any conventional processor, etc., which can realize or execute the disclosed methods, steps, and logic block diagrams in the embodiments of the present application.

[0213] The memory can be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read only memory (PROM), an erasable programmable read only memory (EPROM), an electrically erasable programmable read only memory (EEPROM), etc. The memory is used to store a computer program, and the processor can execute the computer program accordingly after receiving an execution instruction.

[0214] In summary, the application provides an evaluation method and system for anal cleaning effect. The method comprises the following steps: collecting target parameters to obtain simulation monitoring data when a test sample is used to simulate wiping of the anal anatomical structure 10 according to a preset wiping strategy; calling an initial mechanical simulation model, performing finite element simulation based on initial model parameters to obtain comprehensive response data, and comparing the comprehensive response data with the simulation monitoring data to obtain calibration parameters, so as to adjust the skin material parameters in the initial model parameters to obtain a target mechanical simulation model; collecting target parameters to obtain target monitoring data when a to-be-tested sample is used to simulate wiping of the anal anatomical structure 10 according to a target wiping strategy, analyzing the target monitoring data, and obtaining a cleaning effect evaluation report of residual dirt, which realizes calibration of the parameters of the mechanical simulation model during simulation of the anal cleaning process, obtains a more realistic mechanical simulation model, and is used for prediction of the anal cleaning process to evaluate the anal cleaning effect, thereby effectively improving the simulation reality of the anal cleaning process and improving the accuracy of the evaluation of the cleaning effect.

[0215] The above merely illustrates the specific embodiments of the application, but the protection scope of the application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the application, which shall be covered within the protection scope of the application.

Claims

1. A method for evaluating the effectiveness of anal cleaning, characterized in that, include: The control execution component wipes the anal anatomical structure according to the target wiping strategy, and controls the multimodal sensing component to collect the target parameters on the anal anatomical structure to obtain target monitoring data; Based on the amount of residual dirt in the target monitoring data, a cleaning efficiency score is calculated, and a target mechanical simulation model is called to calculate a skin irritation risk score based on the humidity penetration gradient value and pressure value in the target monitoring data, as well as the sample to be tested. The skin irritation risk score is graded according to a preset level and combined with the cleaning efficiency score to form a cleaning effect evaluation report for the sample under test. The step of calling the target mechanical simulation model to calculate the skin irritation risk score based on the humidity penetration gradient value and pressure value in the target monitoring data, as well as the sample to be tested, includes: calling the target mechanical simulation model to analyze the skin moisture balance and the magnitude of chemical irritation based on the humidity penetration gradient value and the component information of the sample to be tested, and assessing the magnitude of mechanical irritation based on the pressure value in the target monitoring data, so as to calculate the skin irritation risk score based on the skin moisture balance, the magnitude of chemical irritation, and the magnitude of mechanical irritation.

2. The method for evaluating the anal cleaning effect according to claim 1, characterized in that, The control execution component wipes the anal anatomical structure of the sample to be tested according to the target wiping strategy, prior to which includes: The execution component is controlled to perform simulated wiping of the anal anatomical structure according to a preset wiping strategy, and the multimodal sensing component is controlled to collect target parameters on the anal anatomical structure to obtain simulated monitoring data. After calling the initial mechanical simulation model and performing finite element simulation based on the initial model parameters to obtain comprehensive response data, the data is compared with the simulated monitoring data to obtain calibration parameters; The skin material parameters in the initial model parameters are adjusted according to the calibration parameters to obtain the target mechanical simulation model.

3. The method for evaluating the anal cleaning effect according to claim 2, characterized in that, After obtaining comprehensive response data by performing finite element simulation based on the initial model parameters using the initial mechanical simulation model, the data is compared with the simulated monitoring data to obtain calibration parameters, including: The initial mechanical simulation model is invoked to perform multiple forward finite element simulations based on the initial model parameters until the number of iterations reaches the preset maximum number of iterations, thereby obtaining the predicted friction force and the predicted equivalent liquid permeability. Based on the pressure distribution changes in the simulated monitoring data and the motion parameters of the actuator, the actual friction force is analyzed and compared with the predicted friction force to obtain the reference coefficients of friction force and elastic modulus. The data on the penetration rate, distribution, and moisture changes of liquid at different skin depths in the simulated monitoring data are analyzed to obtain the actual equivalent liquid permeability of the skin layer. This is then compared with the predicted equivalent liquid permeability to obtain the calibration coefficient of the equivalent liquid permeability. The motion parameters include: velocity, acceleration, joint torque, position, and direction of motion; the target parameters include: pressure value, amount of residual dirt, and humidity penetration gradient value; the skin material parameters include: equivalent liquid permeability of the skin layer, elastic modulus, and coefficient of friction.

4. The method for evaluating the anal cleaning effect according to claim 2, characterized in that, The control execution component wipes the anal anatomical structure according to the target wiping strategy, and prior to this, it also includes: The target mechanical simulation model is invoked to analyze the simulated monitoring data collected under different wiping strategies, and the cleaning effect evaluation value is obtained. A reinforcement learning model is invoked to perform deep learning on the state space of the cleaning effect evaluation value and the simulated monitoring data to obtain reward values ​​for different wiping strategies; wherein, the state space is obtained by the multimodal sensing component monitoring the target parameters; including the surface dirt residue distribution, fold deformation state and internal humidity gradient distribution of the radial folds in the anal anal anatomical structure; The policy parameters of the reward function of the reinforcement learning model are updated according to the reward value to obtain the target wiping policy; The step of calling a reinforcement learning model to perform deep learning on the state space of the cleaning effect evaluation value and the simulated monitoring data to obtain reward values ​​for different wiping strategies includes: designing a reward function that includes cleaning efficiency reward, skin safety reward and resource efficiency reward, and based on the reward function, calling the reinforcement learning algorithm in the reinforcement learning model to perform deep learning on the cleaning effect evaluation value and the state space to calculate the reward value of the current cleaning state-wiping action combination.

5. The method for evaluating the anal cleaning effect according to claim 1, characterized in that, The cleaning efficiency score is calculated based on the amount of residual dirt in the target monitoring data, including: Based on the amount of residual dirt on the radial folds of the anal anatomical structure in the target monitoring data, the depth removal rate, dirt removal rate, and cleaning uniformity are calculated, and then the cleaning efficiency score is obtained.

6. The method for evaluating the anal cleaning effect according to claim 5, characterized in that, The calculation of depth removal rate, dirt removal rate, and cleaning uniformity based on the amount of residual dirt on the radial folds of the anal anatomical structure in the target monitoring data includes: The initial amount of dirt at the bottom of the radial folds is compared with the amount of residual dirt at the corresponding location in the target monitoring data to calculate the depth removal rate. The dirt removal rate is calculated by combining the initial total amount of dirt on the radially folded layer with the residual total amount of dirt at the corresponding location. The standard deviation between the amount of residual dirt in different target areas of the radial folded layer is calculated to obtain the cleaning uniformity. The method involves calling the target mechanical simulation model, analyzing the skin moisture balance and chemical irritation level based on the humidity penetration gradient value in the target monitoring data and the component information of the sample to be tested, and assessing the mechanical irritation level based on the pressure value in the target monitoring data, including: Based on the real-time changes in the humidity permeability gradient value, the transdermal water loss rate and skin moisture content changes are calculated to analyze the skin moisture balance. The magnitude of chemical irritation is predicted based on the humidity penetration gradient value, the pH value of the sample to be tested, and the surfactant content. Based on the real-time changes in the pressure value, the stress distribution on the surface and inside of the radially folded layer is calculated to assess the magnitude of the mechanical irritation.

7. A system for evaluating the effectiveness of anal cleaning, characterized in that, include: The anal anatomy structure, multimodal sensing components, processor, and execution components are connected in sequence. The multimodal sensing component is used to collect target parameters on the anal anatomical structure and obtain target monitoring data when the execution component wipes the sample to be tested on the anal anatomical structure according to the target wiping strategy. The processor is used to calculate a cleaning efficiency score based on the amount of residual dirt in the target monitoring data, and to call the target mechanical simulation model to calculate a skin irritation risk score based on the humidity penetration gradient value and pressure value in the target monitoring data and the sample to be tested. The skin irritation risk score is then classified according to a preset level and combined with the cleaning efficiency score to form a cleaning effect evaluation report for the sample to be tested. The step of calling the target mechanical simulation model to calculate the skin irritation risk score based on the humidity penetration gradient value and pressure value in the target monitoring data, as well as the sample to be tested, includes: calling the target mechanical simulation model to analyze the skin moisture balance and the magnitude of chemical irritation based on the humidity penetration gradient value and the component information of the sample to be tested, and assessing the magnitude of mechanical irritation based on the pressure value in the target monitoring data, so as to calculate the skin irritation risk score based on the skin moisture balance, the magnitude of chemical irritation, and the magnitude of mechanical irritation.

8. The system for evaluating the anal cleaning effect according to claim 7, characterized in that, The multimodal sensing component is also used to collect target parameters on the anal anatomy structure and obtain simulated monitoring data when the processor controls the execution component to perform simulated wiping of the anal anatomy structure with the test sample; The processor is further configured to call the initial mechanical simulation model, perform finite element simulation based on the initial model parameters to obtain comprehensive response data, compare it with the simulation monitoring data to obtain calibration parameters, and adjust the skin material parameters in the initial model parameters to obtain the target mechanical simulation model; The processor is further configured to call a reinforcement learning model to analyze the simulated monitoring data under different wiping strategies, perform deep learning on the obtained cleaning effect evaluation value and the state space in the simulated monitoring data to obtain the reward value of different wiping strategies, and update the policy parameters of the reward function of the reinforcement learning model to obtain the target wiping strategy.

9. The system for evaluating the anal cleaning effect according to claim 7, characterized in that, The anal anatomical structure includes: a simulated anal model connected in pairs, a sphincter drive unit, and radial folds; The sphincter actuation unit includes: a pneumatic artificial muscle and a shape memory alloy wire; The pneumatic artificial muscle and the shape memory alloy wire are evenly distributed around the anus of the simulated anus model to simulate the contraction of the circular sphincter. The radial fold layer includes a supporting base and a hydrogel simulated skin; the radial fold layer is disposed on the supporting base, and the radial fold layer has a ring-shaped radial groove structure to simulate the skin structure around the anus.

10. The system for evaluating the anal cleaning effect according to claim 9, characterized in that, The multimodal sensing components include: a pressure sensing array, an optical detector, and a humidity gradient sensor; The pressure sensing array is embedded in the hydrogel simulated skin in the form of a flexible piezoresistive film, and is used to collect the pressure value of the radial wrinkle layer in real time when the radial wrinkle layer is simulated to be wiped. The optical detector is positioned directly above the radially folded layer and is used to acquire images of residual dirt on the radially folded layer in real time during simulated wiping of the radially folded layer, so as to obtain the amount of residual dirt. The humidity gradient sensor is arranged in layers along the depth of the folds of the hydrogel simulated skin, and is used to collect the humidity penetration gradient value of the liquid on the radial fold layer in real time when the radial fold layer is simulated to wipe.

Images