Clothes sorting management system and method based on artificial intelligence

Abstract

The application discloses a clothes sorting management system and method based on artificial intelligence, and relates to the technical field of artificial intelligence.The system comprises an order information analysis and demand label matching module, a clothes attribute collection and historical data calling module, an adaptive scheme generation module, a sorting priority calculation module, and a sorting execution and data feedback module.The method matches customer demand labels by obtaining order information, collects clothes attributes and calls historical data, cross-analyzes to generate individualized washing adaptive schemes, multi-dimensionally calculates priorities to realize accurate sorting, and synchronously records feedback data.The application realizes accurate adaptation of customer demand and clothes attributes, optimizes sorting resource allocation, improves sorting efficiency and intelligent level, reduces clothes damage rate, and adapts to differentiated demands of multiple types of customers.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, specifically to an artificial intelligence-based clothing sorting management system and method. Background Technology

[0002] In laundry factory garment sorting operations, the sorting process, as the core hub connecting garment collection and washing, directly impacts washing quality, service efficiency, and customer satisfaction. Existing garment sorting management systems and methods are mostly based on the inherent attributes of garments such as material, color, and size, employing standardized sorting rules to process all garments uniformly, without incorporating personalized customer needs into the core considerations of sorting decisions.

[0003] Currently, laundry factories serve a wide range of clients, including hospitals, hotels, businesses, and individuals. These clients have significantly different garment washing needs: hospital linens require high-level sterilization and residue-free treatment, hotel linens need to be ironed crisp, and valuable personal garments require meticulous care to prevent damage. However, existing systems lack targeted correlation analysis between customer needs and garment attributes. Simply sorting according to uniform standards often leads to a mismatch between needs and services, such as hospital linens failing to meet sterilization requirements or valuable garments being damaged by conventional washing processes. Summary of the Invention

[0004] The purpose of this invention is to provide an artificial intelligence-based clothing sorting management system and method to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: An AI-based clothing sorting and management method includes the following steps: S1. Obtain the order information corresponding to the clothes to be sorted. The order information includes customer type identifier, order service demand coefficient, preset delivery time limit and service demand terms. Based on the customer type identifier, associate the preset customer type and obtain the customer demand tag that is preset to be bound to the corresponding customer type. The demand tag includes functional constraint threshold, care level standard, sorting isolation requirements and packaging specifications. S2. Collect the attribute information of the clothes to be sorted. The attribute information includes the type of clothing material, the classification of uses, and historical washing records. The historical washing records include the previous adapted process parameters, damage feedback data, and compatibility scores. S3. Cross-analyze customer demand tags and clothing attribute information, calculate the compatibility score between each functional constraint threshold in the demand tags and clothing attributes, and generate corresponding clothing washing adaptation schemes based on the compatibility score. The adaptation schemes include process parameter ranges, special reagent types and equipment functional requirements. S4. Determine the service level weight based on the order service demand coefficient in the order information, and calculate the order urgency coefficient; at the same time, calculate the demand complexity coefficient of the washing adaptation solution, and obtain the corresponding priority score, and divide the sorting priority according to the priority score; S5. Based on sorting priority and washing adaptation scheme, control the sorting equipment to transport the clothes to be sorted to the dedicated sorting channel and target workstation that match the functional requirements of the equipment, and simultaneously record the correlation data of clothing attributes, demand tags, priority scores and sorting paths, and feed back the sorting completion status and channel matching results to the system in real time.

[0006] Furthermore, S1 includes the following: The system interface obtains the order information corresponding to the clothes to be sorted, and extracts the customer type identifier, order service demand coefficient, preset delivery time limit and service demand terms from the order information; the customer type identifier is a preset unique code, the order service demand coefficient is a quantitative value of 1-5 levels, and the service demand terms include quantitative indicators related to functional constraint thresholds. A mapping database is constructed between customer type identifiers and preset customer types. This database stores a one-to-one correspondence between preset codes and preset customer types. The extracted customer type identifiers are precisely matched with the preset codes in the mapping database. If a match is successful, the corresponding preset customer type is directly determined. If a match fails, core functional keywords from the service requirement terms are extracted using a preset keyword dictionary. This dictionary contains typical functional requirement keywords corresponding to each preset customer type. The extraction method for core functional keywords is as follows: based on the preset keyword dictionary, a text segmentation algorithm is used to extract relevant keywords from the service requirement terms, and redundant stop words are removed to obtain a set of core functional keywords. The extracted core functional keywords are then matched with the typical requirement keywords of each preset customer type for similarity. The similarity calculation formula is: S = (M / N) × 100%, where M is the number of successfully matched core keywords, and N is the total number of core keywords in the service requirement terms. If S ≥ 80%, the corresponding preset customer type is determined. If S < 80%, a customer type matching failure message is generated and fed back to the system in real time. The order sorting process is paused until manual verification is performed, and the matching step is re-executed. The service level base weight Wb is calculated based on the order service demand coefficient, and Wb = L / Lmax, where L is the order service demand coefficient, L∈[1,5], and Lmax is the maximum service level parameter and Lmax=5; a predefined customer type and customer demand tag association database is constructed. Each predefined customer type in the association database corresponds to a set of weighted demand tags. The sum of the weights of each demand tag is 1. The weight of core demand tags is ≥60%, and the weight of non-core demand tags is ≤40%; according to the determined predefined customer type, the corresponding demand tag group is retrieved from the association database as the customer demand tag.

[0007] Furthermore, S2 includes the following: By collecting basic attribute information and garment label images of the clothes to be sorted, and combining this with the material descriptions provided by customers in the order information, the material type and usage classification of the clothes are jointly determined. The material type is determined through a triple verification process: extracting visual features of the garment's fiber texture and surface gloss, identifying the preset material identifier in the label image, verifying consistency with the order's material description, and encoding it according to preset classification rules from M1 to Mn, where n is the total number of material types. Usage classification is determined by identifying structural features of the garment's collar style, pocket design, and size specifications, matching them with keywords used in the scene description in the order, and encoding it according to preset classification rules from U1 to Um, where m is the total number of usage classifications. The classification by application is achieved through the following steps: The system performs contour segmentation, edge detection, and region labeling on clothing to extract structural features such as neckline style, number of pockets, size specifications, and overall contour shape. These extracted features are then converted into structured feature items for neckline type, pocket presence / absence, size range, and contour shape. The system also segments the scenario description text used in orders, removing stop words and extracting usage keywords such as bedding, tops, trousers, hotel linens, and business formal wear. A mapping rule base between structural features and usage keywords is established. The extracted structured feature items are matched against this rule base item by item, and the number of successfully matched feature items is counted. The feature matching degree is calculated, and when the feature matching degree exceeds a preset threshold, the usage classification is completed. The unique garment IDs are extracted from the obtained order information, and a database linking these unique garment IDs with historical washing records is constructed. This database stores the previous adaptation process parameters, damage feedback data, and single-time compatibility scores corresponding to each garment ID. The extracted garment IDs are then precisely matched with the database. If a match is successful, all corresponding historical washing records are retrieved. If a match fails, a multi-dimensional feature combination of the garment is extracted using a visual acquisition device and compared with the feature combinations of historical garments in the database. The similarity calculation formula is Sc = (P / Q) × 100%, where P is the number of successfully matched feature items and Q is the total number of feature items in the multi-dimensional feature combination. When Sc ≥ 95%, the corresponding historical washing record is retrieved. The clothing compatibility score Pa is calculated based on the retrieved historical washing records. The calculation formula is: Pa=α×A+β×(1-B), where α is the process compatibility weight, β is the damage impact weight, and α+β=1. The specific values ​​are determined according to the clothing material type and customer demand labels. A is the average compatibility between the previous compatible process and the clothing material, and A=∑Ai / k, k is the number of historical washes, Ai is the compatibility score of the i-th compatible process, Ai=γ×C+δ×D, where γ is the process parameter matching weight, δ is the washing effect weight, and γ+δ=1. The specific values ​​are determined according to the clothing usage classification and the functional constraint threshold in the customer demand labels. C represents the matching degree of process parameters, and C = (number of parameters in the adapted process that meet the requirements of clothing material / total number of process parameters) × 100%, C ∈ [0,1]; D represents the washing effect compliance rate, D = (total area of ​​stains on clothing before washing - total area of ​​stains on clothing after washing) / total area of ​​stains on clothing before washing, D ∈ [0,1]; B represents the average proportion of damaged area in historical washing, B = ∑Bi / k, Bi is the proportion of damaged area in the i-th washing, which is equal to the ratio of the total damaged area of ​​clothing after washing to the total surface area of ​​clothing, and the total damaged area of ​​clothing includes the quantified area corresponding to holes, snags, and loose threads, all of which are obtained by segmentation and measurement through a visual acquisition device, Bi ∈ [0,1]; By integrating material type coding, usage classification coding, compatibility comprehensive score Pa, previous compatibility process parameters, and average damage area percentage B, a complete dataset of clothing attribute information is formed.

[0008] Furthermore, S3 includes the following: The customer demand tags and clothing attribute information are preprocessed in a structured manner to extract the determined customer demand tag groups and clarify the target weight ω of each demand tag. For example, the value range is 0.1-0.8, the sum of the weights is 1, the core demand tags ω≥0.3, and the non-core demand tags ω≤0.2. The target weights are derived from the preset weights of the S1 association database. The resulting clothing attribute information dataset is extracted, and the core attributes associated with the demand tags are screened, specifically including: material type code M, usage classification code U, comprehensive suitability score Pa, average damaged area ratio B, and the set of previous suitable process parameters. A cross-analysis dimension mapping table is established to correspond the functional constraint thresholds in the demand tags, such as washing temperature threshold T0, moisture content threshold H0, and detergent type restriction D0, to the core attributes of the clothing, such as material M, usage U, Pa, and B, forming a mapping relationship between functional constraint types and clothing attribute dimensions. For each functional constraint threshold, combined with its corresponding clothing attribute dimension, a single fit score Sf for the corresponding clothing attribute is calculated. The specific formula is: Sf=ω×θ×[φ×F(M,U)+σ×G(Pa,B)], where θ is the functional constraint type coefficient, for example, washing temperature constraint θ=0.9, moisture content constraint θ=0.85, detergent type constraint θ=0.8, other functional constraints θ=0.75, a preset fixed value; φ is the material and usage fit coefficient, determined according to the combination relationship between material type M and usage category U, for example, when M=wool material M3, U=business formal wear U2, φ=1.15, when M=cotton material M1, U=casual wear U5, φ=1.0; F(M,U) F(M,U) is the basic adaptation value for material and application. It is calculated as follows: F(M,U) = (baseline score of material M adapting to the constraint + baseline score of application U adapting to the constraint) / 2. The baseline score is preset to 0-1 points. For example, the baseline score of wool material adapting to the "temperature ≤30℃" constraint is 0.95, and the baseline score of cotton material adapting to the constraint is 0.8. σ is the historical data correction coefficient, which is dynamically adjusted according to the comprehensive adaptation score Pa and the average proportion of damaged area B: σ = 1 + 0.05 × (Pa - 0.5) - 0.1 × B. If σ < 0.9, then take 0.9; if σ > 1.1, then take 1.1. G(Pa,B) is the historical data adaptation value. G(Pa,B) = Pa × (1 - B), and G(Pa,B) ∈ [0,1]. Calculate the overall fit score St and determine the fit level. The formula for calculating the overall fit score St is: St=∑(Sfi×ωi) / ∑ωi, where Sfi is the fit score of the i-th functional constraint, ωi is the label weight corresponding to the i-th functional constraint, and St∈[0,1]. When St≥0.8, it is marked as Level 1 fit; when 0.6≤St<0.8, it is marked as Level 2 fit; when St<0.6, it is marked as Level 3 fit; if the Sfi corresponding to the core requirement label is <0.5, the fit level is directly reduced to Level 3. Based on the adaptation level and clothing attributes, a clothing washing adaptation plan is generated. For Level 1 adaptation, the previously retrieved adaptation process parameters are used as the benchmark, with a fluctuation range of ≤±5%. For example, if the previous washing temperature was 30℃, the range is 28.5-31.5℃. For Level 2 adaptation, the standard process parameters corresponding to material M are used as the benchmark, combined with functional constraint threshold adjustments, with a fluctuation range of ≤±10%, and the constraint thresholds of the core requirement label must be met. For example, if the washing temperature threshold is ≤30℃, the range is 25-30℃. For Level 3 adaptation, the functional constraint thresholds of the core requirement label are prioritized, with a fluctuation range of ≤±15% for process parameters, while reducing the washing intensity, for example, the spin speed is ≤600r. / min, washing time reduced by 20%; if the average damaged area ratio B ≥ 0.1, the process parameter range is further relaxed by 5% based on the corresponding adaptation level; according to the material M, functional constraint threshold and adaptation level selection, the first-level adaptation uses the previously adapted reagents, the second-level adaptation uses general-purpose special reagents, and the third-level adaptation uses mild special reagents, while avoiding incompatible reagents associated with historical damage feedback data; according to the process parameter range, material M and sorting isolation requirements, the functions that the equipment must have are clarified, such as the first-level adaptation requires the equipment to support precise temperature control ±0.5℃, the third-level adaptation requires the equipment to support damage protection function, and when the sorting isolation requirement is "independent channel", the equipment must support single-channel independent operation; Integrate process parameter ranges, special reagent types, and equipment functional requirements to form a structured adaptation solution document, associate it with the unique garment number and total compatibility score St, and store it in the system database.

[0009] Furthermore, S4 includes the following: Extract the order service demand coefficient L and the preset delivery time limit T0. Extract the core parameters from the generated washing adaptation plan. At the same time, obtain the current system time T1 and the order creation time Ts. Calculate the order elapsed time Tu, and Tu = T1 - Ts. Based on the calculated service level base weight Wb, introduce the customer type correction coefficient λ to calculate the service level weight Wd, and Wd = Wb × λ, where the customer type correction coefficient is set according to the preset customer type. The urgency level is quantified based on the difference between the preset delivery time limit T0 and the order elapsed time Tu. The order urgency coefficient We is calculated, and We = 1 / [(T0-Tu)+1]×100%; if T0-Tu≤0, then We =100%. The complexity coefficient Wc of sorting requirements is calculated based on the core parameters of the washing adaptation scheme. The calculation formula is: Wc=(a×Cn+b×Rn+c×En) / (Cn_max+Rn_max+En_max), where a, b, and c are weight coefficients, corresponding to the influence weights of process parameters, reagent types, and equipment functions, respectively, and a+b+c=1; Cn is the number of process parameters, Rn is the number of special reagent types, En is the number of equipment function requirements, Cn_max is the preset maximum number of process parameters, Rn_max is the preset maximum number of special reagent types, and En_max is the preset maximum number of equipment function requirements. Calculate the priority score Sp, where Sp = (d × Wd + e × We + f × Wc) × K, where d, e, and f are the priority dimension weights, corresponding to the influence weights of service level, order urgency, and demand complexity, respectively, and d + e + f = 1; K is the core demand adaptation correction item, determined based on the individual adaptation score Sfi of the core demand tag. For example, if the Sfi of all core demand tags is ≥ 0.6, then K = 1.1; if there is one core demand tag with Sfi < 0.6, then K = 0.9; if there are two or more core demand tags with Sfi < 0.6, then K = 0.7; obtain the value range of the priority score Sp, and divide it equally according to the number M of preset sorting priorities to obtain the value interval corresponding to each preset sorting priority; compare the calculated priority score Sp with the value interval corresponding to each preset sorting priority, output the sorting priority level, and store it in the system along with the unique clothing number.

[0010] An AI-based clothing sorting and management system includes: an order information parsing and demand tag matching module, a clothing attribute collection and historical data retrieval module, an adaptation scheme generation module, a sorting priority calculation module, and a sorting execution and data feedback module. The order information parsing and demand tag matching module is used to obtain the order information of the clothes to be sorted, parse and extract key parameters and associate them with preset customer types, and retrieve the corresponding customer demand tags. The clothing attribute collection and historical data retrieval module is used to collect basic attribute information and washing label images of clothing, determine the material type and usage classification, retrieve the corresponding historical washing records, and integrate them to form an attribute information dataset. The adaptation solution generation module is used to cross-analyze customer demand tags and clothing attribute information, calculate the adaptation score, and generate a washing adaptation solution that includes process parameter range, special reagent type, and equipment functional requirements. The sorting priority calculation module is used to calculate the service level weight, order urgency coefficient, and demand complexity coefficient based on order information and washing adaptation plan, obtain priority score, and divide sorting priority. The sorting execution and data feedback module is used to control the operation of sorting equipment based on sorting priority and adaptation scheme, record related data, and provide real-time feedback on sorting status and channel matching results.

[0011] Furthermore, the order information parsing and demand tag matching module includes an order information acquisition unit and a demand tag matching unit; The order information acquisition unit obtains the order information corresponding to the clothes to be sorted through the system interface, and parses and extracts the customer type identifier, order service demand coefficient, preset delivery time limit and service demand terms to complete the validity verification of the order information; the demand tag matching unit matches the customer type identifier with the preset customer type, determines the customer type through precise matching in the mapping database or keyword similarity matching, and retrieves the weighted demand tag group that is preset to be bound to the customer type.

[0012] Furthermore, the clothing attribute collection and historical data retrieval module includes a clothing attribute collection unit and a historical data retrieval unit; The clothing attribute acquisition unit collects basic attribute information and washing label images of clothing through a visual acquisition device. Combined with the material description filled in by the customer in the order, it determines the type of clothing material through triple verification and classifies the use by matching structural features and keywords. The historical data retrieval unit extracts the unique number of clothing from the order information, retrieves historical washing records through precise matching or multi-dimensional feature combination similarity matching, calculates the comprehensive score of clothing suitability, and integrates them to form a complete clothing attribute information dataset.

[0013] Furthermore, the adaptation scheme generation module includes a cross-analysis unit and an adaptation scheme output unit; The cross-analysis unit performs structured preprocessing on customer demand tags and clothing attribute information, establishes cross-analysis dimension mapping relationships, calculates the individual and total fit scores of each functional constraint threshold and clothing attribute, and determines the fit level. The fit solution output unit determines the process parameter range, special reagent type and equipment function requirements based on the fit level, clothing material type and historical damage status, avoids prohibited reagents, integrates and forms a structured clothing washing fit solution and stores it.

[0014] Furthermore, the sorting priority calculation module includes a priority parameter calculation unit and a priority division unit, and the sorting execution and data feedback module includes a sorting control unit and a data recording feedback unit. The priority parameter calculation unit extracts the core parameters from the order and adaptation plan, calculates the service level weight, order urgency coefficient and requirement complexity coefficient, and obtains the priority score by combining the core requirement adaptation correction items; the priority division unit obtains the value range of the priority score, divides the interval equally according to the preset sorting priority quantity, and determines the sorting priority level by score comparison; Based on sorting priorities and washing adaptation schemes, the sorting control unit controls the sorting equipment to transport clothes to the matching dedicated sorting channels and target workstations; the data recording and feedback unit synchronously records the correlation data of clothing attributes, demand tags, priority scores and sorting paths, and feeds back the sorting completion status and channel matching results to the system in real time.

[0015] Compared with existing technologies, the beneficial effects of this invention are as follows: Through a dual-matching mechanism based on customer type and a weighted demand tag design, core service needs are accurately identified. Resource allocation is optimized based on service levels to adapt to the needs of multiple customer types, ensuring a high-value customer service experience. Triple verification is used to determine materials, and dual matching retrieves historical data, reducing the probability of attribute misjudgment and providing comprehensive and reliable data support for subsequent adaptation solution generation. Based on compatibility scores, differentiating process and reagent solutions are generated, and parameters are dynamically adjusted based on historical damage data to reduce washing errors and rework. Priority is calculated and refined based on multiple dimensions to ensure that urgent orders are processed first, avoiding resource idleness or congestion and shortening delivery cycles. Data correlation and traceability are achieved throughout the entire sorting process, providing data basis for system optimization, promoting the transformation of management from experience-driven to data-driven, and improving standardization. Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a clothing sorting and management system based on artificial intelligence according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 The present invention provides the following technical solution: An AI-based clothing sorting and management system includes: an order information parsing and demand tag matching module, a clothing attribute collection and historical data retrieval module, an adaptation scheme generation module, a sorting priority calculation module, and a sorting execution and data feedback module. The order information parsing and demand tag matching module is used to obtain the order information of the clothes to be sorted, parse and extract key parameters and associate them with preset customer types, and retrieve the corresponding customer demand tags. The clothing attribute collection and historical data retrieval module is used to collect basic attribute information and washing label images of clothing, determine the material type and usage classification, retrieve the corresponding historical washing records, and integrate them to form an attribute information dataset. The adaptation solution generation module is used to cross-analyze customer demand tags and clothing attribute information, calculate the adaptation score, and generate a washing adaptation solution that includes process parameter range, special reagent type, and equipment functional requirements. The sorting priority calculation module is used to calculate the service level weight, order urgency coefficient, and demand complexity coefficient based on order information and washing adaptation plan, obtain priority score, and divide sorting priority. The sorting execution and data feedback module is used to control the operation of sorting equipment based on sorting priority and adaptation scheme, record related data, and provide real-time feedback on sorting status and channel matching results.

[0019] The order information parsing and demand tag matching module includes an order information acquisition unit and a demand tag matching unit; The order information acquisition unit obtains the order information corresponding to the clothes to be sorted through the system interface, and parses and extracts the customer type identifier, order service demand coefficient, preset delivery time limit and service demand terms to complete the validity verification of the order information; the demand tag matching unit matches the customer type identifier with the preset customer type, determines the customer type through precise matching in the mapping database or keyword similarity matching, and retrieves the weighted demand tag group that is preset to be bound to the customer type.

[0020] The clothing attribute acquisition and historical data retrieval module includes a clothing attribute acquisition unit and a historical data retrieval unit; The clothing attribute acquisition unit collects basic attribute information and washing label images of clothing through a visual acquisition device. Combined with the material description filled in by the customer in the order, it determines the type of clothing material through triple verification and classifies the use by matching structural features and keywords. The historical data retrieval unit extracts the unique number of clothing from the order information, retrieves historical washing records through precise matching or multi-dimensional feature combination similarity matching, calculates the comprehensive score of clothing suitability, and integrates them to form a complete clothing attribute information dataset.

[0021] The adaptation scheme generation module includes a cross-analysis unit and an adaptation scheme output unit; The cross-analysis unit performs structured preprocessing on customer demand tags and clothing attribute information, establishes cross-analysis dimension mapping relationships, calculates the individual and total fit scores of each functional constraint threshold and clothing attribute, and determines the fit level. The fit solution output unit determines the process parameter range, special reagent type and equipment function requirements based on the fit level, clothing material type and historical damage status, avoids prohibited reagents, integrates and forms a structured clothing washing fit solution and stores it.

[0022] The sorting priority calculation module includes a priority parameter calculation unit and a priority division unit; the sorting execution and data feedback module includes a sorting control unit and a data recording feedback unit. The priority parameter calculation unit extracts the core parameters from the order and adaptation plan, calculates the service level weight, order urgency coefficient and requirement complexity coefficient, and obtains the priority score by combining the core requirement adaptation correction items; the priority division unit obtains the value range of the priority score, divides the interval equally according to the preset sorting priority quantity, and determines the sorting priority level by score comparison; Based on sorting priorities and washing adaptation schemes, the sorting control unit controls the sorting equipment to transport clothes to the matching dedicated sorting channels and target workstations; the data recording and feedback unit synchronously records the correlation data of clothing attributes, demand tags, priority scores and sorting paths, and feeds back the sorting completion status and channel matching results to the system in real time.

[0023] An AI-based clothing sorting and management method includes the following steps: S1. Obtain the order information corresponding to the clothes to be sorted. The order information includes customer type identifier, order service demand coefficient, preset delivery time limit and service demand terms. Based on the customer type identifier, associate the preset customer type and obtain the customer demand tag that is preset to be bound to the corresponding customer type. The demand tag includes functional constraint threshold, care level standard, sorting isolation requirements and packaging specifications. S2. Collect the attribute information of the clothes to be sorted. The attribute information includes the type of clothing material, the classification of uses, and historical washing records. The historical washing records include the previous adapted process parameters, damage feedback data, and compatibility scores. S3. Cross-analyze customer demand tags and clothing attribute information, calculate the compatibility score between each functional constraint threshold in the demand tags and clothing attributes, and generate corresponding clothing washing adaptation schemes based on the compatibility score. The adaptation schemes include process parameter ranges, special reagent types and equipment functional requirements. S4. Determine the service level weight based on the order service demand coefficient in the order information, and calculate the order urgency coefficient; at the same time, calculate the demand complexity coefficient of the washing adaptation solution, and obtain the corresponding priority score, and divide the sorting priority according to the priority score; S5. Based on sorting priority and washing adaptation scheme, control the sorting equipment to transport the clothes to be sorted to the dedicated sorting channel and target workstation that match the functional requirements of the equipment, and simultaneously record the correlation data of clothing attributes, demand tags, priority scores and sorting paths, and feed back the sorting completion status and channel matching results to the system in real time.

[0024] S1 includes the following: The system interface obtains the order information corresponding to the clothes to be sorted, and extracts the customer type identifier, order service demand coefficient, preset delivery time limit and service demand terms from the order information; the customer type identifier is a preset unique code, the order service demand coefficient is a quantitative value of 1-5 levels, and the service demand terms include quantitative indicators related to functional constraint thresholds. A mapping database is constructed between customer type identifiers and preset customer types. This database stores a one-to-one correspondence between preset codes and preset customer types. The extracted customer type identifiers are precisely matched with the preset codes in the mapping database. If a match is successful, the corresponding preset customer type is directly determined. If a match fails, core functional keywords from the service requirement terms are extracted using a preset keyword dictionary. This dictionary contains typical functional requirement keywords corresponding to each preset customer type. The extraction method for core functional keywords is as follows: based on the preset keyword dictionary, a text segmentation algorithm is used to extract relevant keywords from the service requirement terms, and redundant stop words are removed to obtain a set of core functional keywords. The extracted core functional keywords are then matched with the typical requirement keywords of each preset customer type for similarity. The similarity calculation formula is: S = (M / N) × 100%, where M is the number of successfully matched core keywords, and N is the total number of core keywords in the service requirement terms. If S ≥ 80%, the corresponding preset customer type is determined. If S < 80%, a customer type matching failure message is generated and fed back to the system in real time. The order sorting process is paused until manual verification is performed, and the matching step is re-executed. The service level base weight Wb is calculated based on the order service demand coefficient, and Wb = L / Lmax, where L is the order service demand coefficient, L∈[1,5], and Lmax is the maximum service level parameter and Lmax=5; a predefined customer type and customer demand tag association database is constructed. Each predefined customer type in the association database corresponds to a set of weighted demand tags. The sum of the weights of each demand tag is 1. The weight of core demand tags is ≥60%, and the weight of non-core demand tags is ≤40%; according to the determined predefined customer type, the corresponding demand tag group is retrieved from the association database as the customer demand tag.

[0025] In this embodiment, it is assumed that the system obtains the order information of the clothes to be sorted through the order interface, and extracts the following information: customer type identifier is CT-003, order service demand coefficient L=4, preset delivery time limit is 24 hours, and service demand terms are "high temperature disinfection of linens ≥99.9%, ironing and crisping, individual packaging". Assuming the system has built a mapping database, where the preset code CT-003 has a one-to-one correspondence with "hotel customers", the extracted CT-003 is precisely matched with the mapping database. If the match is successful, the preset customer type is directly determined to be a hotel customer. The basic weight of the service level is calculated using the formula Wb=L / Lmax. Substituting L=4 and Lmax=5, we get Wb=4 / 5=0.8. The corresponding requirement tag groups for hotel customers in the associated database are: The core requirement tags are high-temperature disinfection threshold (weight 35%) and rigid care standards (weight 35%). Non-core requirement tags are individual packaging specifications (weight 15%) and sorting and isolation requirements (weight 15%). The sum of the weights of all tags is 1, and the core tag weight accounts for 70% ≥ 60%. The system retrieves this tag group as the customer requirement tags.

[0026] S2 includes the following: By collecting basic attribute information and garment label images of the clothes to be sorted, and combining this with the material descriptions provided by customers in the order information, the material type and usage classification of the clothes are jointly determined. The material type is determined through a triple verification process: extracting visual features of the garment's fiber texture and surface gloss, identifying the preset material identifier in the label image, verifying consistency with the order's material description, and encoding it according to preset classification rules from M1 to Mn, where n is the total number of material types. Usage classification is determined by identifying structural features of the garment's collar style, pocket design, and size specifications, matching them with keywords used in the scene description in the order, and encoding it according to preset classification rules from U1 to Um, where m is the total number of usage classifications. The classification by application is achieved through the following steps: The system performs contour segmentation, edge detection, and region labeling on clothing to extract structural features such as neckline style, number of pockets, size specifications, and overall contour shape. These extracted features are then converted into structured feature items for neckline type, pocket presence / absence, size range, and contour shape. The system also segments the scenario description text used in orders, removing stop words and extracting usage keywords such as bedding, tops, trousers, hotel linens, and business formal wear. A mapping rule base between structural features and usage keywords is established. The extracted structured feature items are matched against this rule base item by item, and the number of successfully matched feature items is counted. The feature matching degree is calculated, and when the feature matching degree exceeds a preset threshold, the usage classification is completed. In this embodiment, the clothing to be sorted is subjected to contour segmentation, edge detection and region marking, and the structural features are extracted as follows: no collar style, no pocket design, size specification of 180cm×220cm, and overall rectangular outline shape. The above structural features are converted into structured feature items: collar type = no collar, pocket presence or absence = none, size range = extra large size, outline shape = rectangle. The text describing the scenario in the order, "hotel room bedding, bulk washing and changing," was segmented into words. After removing stop words, the usage keywords were extracted as: hotel, linens, bedding. A pre-established mapping rule library for structural features and usage keywords is included, which contains the rule: collarless + no pockets + oversized + rectangular outline → bedding / hotel linens; The above structured features are matched item by item with the mapping rule base. The number of features matched successfully this time is 4, and the total number of structured features is 4. The feature matching degree is calculated by formula = 4 / 4 × 100% = 100%. Assuming the preset threshold is 85%, the feature matching degree is greater than the preset threshold, and the use classification is completed. It is coded as U3 (linen) according to the preset classification rule.

[0027] The unique garment IDs are extracted from the obtained order information, and a database linking these unique garment IDs with historical washing records is constructed. This database stores the previous adaptation process parameters, damage feedback data, and single-time compatibility scores corresponding to each garment ID. The extracted garment IDs are then precisely matched with the database. If a match is successful, all corresponding historical washing records are retrieved. If a match fails, a multi-dimensional feature combination of the garment is extracted using a visual acquisition device and compared with the feature combinations of historical garments in the database. The similarity calculation formula is Sc = (P / Q) × 100%, where P is the number of successfully matched feature items and Q is the total number of feature items in the multi-dimensional feature combination. When Sc ≥ 95%, the corresponding historical washing record is retrieved. The clothing compatibility score Pa is calculated based on the retrieved historical washing records. The calculation formula is: Pa=α×A+β×(1-B), where α is the process compatibility weight, β is the damage impact weight, and α+β=1. The specific values ​​are determined according to the clothing material type and customer demand labels. A is the average compatibility between the previous compatible process and the clothing material, and A=∑Ai / k, k is the number of historical washes, Ai is the compatibility score of the i-th compatible process, Ai=γ×C+δ×D, where γ is the process parameter matching weight, δ is the washing effect weight, and γ+δ=1. The specific values ​​are determined according to the clothing usage classification and the functional constraint threshold in the customer demand labels. C represents the matching degree of process parameters, and C = (number of parameters in the adapted process that meet the requirements of clothing material / total number of process parameters) × 100%, C ∈ [0,1]; D represents the washing effect compliance rate, D = (total area of ​​stains on clothing before washing - total area of ​​stains on clothing after washing) / total area of ​​stains on clothing before washing, D ∈ [0,1]; B represents the average proportion of damaged area in historical washing, B = ∑Bi / k, Bi is the proportion of damaged area in the i-th washing, which is equal to the ratio of the total damaged area of ​​clothing after washing to the total surface area of ​​clothing, and the total damaged area of ​​clothing includes the quantified area corresponding to holes, snags, and loose threads, all of which are obtained by segmentation and measurement through a visual acquisition device, Bi ∈ [0,1]; By integrating material type coding, usage classification coding, compatibility comprehensive score Pa, previous compatibility process parameters, and average damage area percentage B, a complete dataset of clothing attribute information is formed.

[0028] S3 includes the following: The customer demand tags and clothing attribute information are preprocessed in a structured manner to extract the determined customer demand tag groups and clarify the target weight ω of each demand tag. For example, the value range is 0.1-0.8, the sum of the weights is 1, the core demand tags ω≥0.3, and the non-core demand tags ω≤0.2. The target weights are derived from the preset weights of the S1 association database. The resulting clothing attribute information dataset is extracted, and the core attributes associated with the demand tags are screened, specifically including: material type code M, usage classification code U, comprehensive suitability score Pa, average damaged area ratio B, and the set of previous suitable process parameters. A cross-analysis dimension mapping table is established to correspond the functional constraint thresholds in the demand tags, such as washing temperature threshold T0, moisture content threshold H0, and detergent type restriction D0, to the core attributes of the clothing, such as material M, usage U, Pa, and B, forming a mapping relationship between functional constraint types and clothing attribute dimensions. For each functional constraint threshold, combined with its corresponding clothing attribute dimension, a single fit score Sf for the corresponding clothing attribute is calculated. The specific formula is: Sf=ω×θ×[φ×F(M,U)+σ×G(Pa,B)], where θ is the functional constraint type coefficient, for example, washing temperature constraint θ=0.9, moisture content constraint θ=0.85, detergent type constraint θ=0.8, other functional constraints θ=0.75, a preset fixed value; φ is the material and usage fit coefficient, determined according to the combination relationship between material type M and usage category U, for example, when M=wool material M3, U=business formal wear U2, φ=1.15, when M=cotton material M1, U=casual wear U5, φ=1.0; F(M,U) F(M,U) is the basic adaptation value for material and application. It is calculated as follows: F(M,U) = (baseline score of material M adapting to the constraint + baseline score of application U adapting to the constraint) / 2. The baseline score is preset to 0-1 points. For example, the baseline score of wool material adapting to the "temperature ≤30℃" constraint is 0.95, and the baseline score of cotton material adapting to the constraint is 0.8. σ is the historical data correction coefficient, which is dynamically adjusted according to the comprehensive adaptation score Pa and the average proportion of damaged area B: σ = 1 + 0.05 × (Pa - 0.5) - 0.1 × B. If σ < 0.9, then take 0.9; if σ > 1.1, then take 1.1. G(Pa,B) is the historical data adaptation value. G(Pa,B) = Pa × (1 - B), and G(Pa,B) ∈ [0,1]. Calculate the overall fit score St and determine the fit level. The formula for calculating the overall fit score St is: St=∑(Sfi×ωi) / ∑ωi, where Sfi is the fit score of the i-th functional constraint, ωi is the label weight corresponding to the i-th functional constraint, and St∈[0,1]. When St≥0.8, it is marked as Level 1 fit; when 0.6≤St<0.8, it is marked as Level 2 fit; when St<0.6, it is marked as Level 3 fit; if the Sfi corresponding to the core requirement label is <0.5, the fit level is directly reduced to Level 3. Based on the adaptation level and clothing attributes, a clothing washing adaptation plan is generated. For Level 1 adaptation, the previously retrieved adaptation process parameters are used as the benchmark, with a fluctuation range of ≤±5%. For example, if the previous washing temperature was 30℃, the range is 28.5-31.5℃. For Level 2 adaptation, the standard process parameters corresponding to material M are used as the benchmark, combined with functional constraint threshold adjustments, with a fluctuation range of ≤±10%, and the constraint thresholds of the core requirement label must be met. For example, if the washing temperature threshold is ≤30℃, the range is 25-30℃. For Level 3 adaptation, the functional constraint thresholds of the core requirement label are prioritized, with a fluctuation range of ≤±15% for process parameters, while reducing the washing intensity, for example, the spin speed is ≤600r. / min, washing time reduced by 20%; if the average damaged area ratio B ≥ 0.1, the process parameter range is further relaxed by 5% based on the corresponding adaptation level; according to the material M, functional constraint threshold and adaptation level selection, the first-level adaptation uses the previously adapted reagents, the second-level adaptation uses general-purpose special reagents, and the third-level adaptation uses mild special reagents, while avoiding incompatible reagents associated with historical damage feedback data; according to the process parameter range, material M and sorting isolation requirements, the functions that the equipment must have are clarified, such as the first-level adaptation requires the equipment to support precise temperature control ±0.5℃, the third-level adaptation requires the equipment to support damage protection function, and when the sorting isolation requirement is "independent channel", the equipment must support single-channel independent operation; Integrate process parameter ranges, special reagent types, and equipment functional requirements to form a structured adaptation solution document, associate it with the unique garment number and total compatibility score St, and store it in the system database.

[0029] In this embodiment, an example of the material-application benchmark score is as follows: Functional constraint: Washing temperature ≤30℃; Wool material + business formal wear: Material baseline score 0.95, usage baseline score 0.93, material-usage basic fit value F(M,U)=(0.95+0.93) / 2=0.94; Cotton material + casual wear: Material baseline score 0.80, usage baseline score 0.82, F(M,U)=(0.80+0.82) / 2=0.81; Examples of standard process parameters are as follows: Wool material (code M3): Standard washing temperature 30℃, standard spin speed 550r / min, standard washing time 18min; Cotton material (code M2): Standard washing temperature 40℃, standard spin speed 800r / min, standard washing time 25min; The above examples illustrate common materials, uses, and core functional constraints in clothing sorting. Technical personnel can extend other combinations based on this logic. Furthermore, the benchmark scores and process parameters in the above examples can be fine-tuned according to actual business scenarios, with an adjustment range not exceeding ±10%, without affecting the core logic of this technical solution.

[0030] S4 includes the following: Extract the order service demand coefficient L and the preset delivery time limit T0. Extract the core parameters from the generated washing adaptation plan. At the same time, obtain the current system time T1 and the order creation time Ts. Calculate the order elapsed time Tu, and Tu = T1 - Ts. Based on the calculated service level base weight Wb, introduce the customer type correction coefficient λ to calculate the service level weight Wd, and Wd = Wb × λ, where the customer type correction coefficient is set according to the preset customer type. The urgency level is quantified based on the difference between the preset delivery time limit T0 and the order elapsed time Tu. The order urgency coefficient We is calculated, and We = 1 / [(T0-Tu)+1]×100%; if T0-Tu≤0, then We =100%. The complexity coefficient Wc of sorting requirements is calculated based on the core parameters of the washing adaptation scheme. The calculation formula is: Wc=(a×Cn+b×Rn+c×En) / (Cn_max+Rn_max+En_max), where a, b, and c are weight coefficients, corresponding to the influence weights of process parameters, reagent types, and equipment functions, respectively, and a+b+c=1; Cn is the number of process parameters, Rn is the number of special reagent types, En is the number of equipment function requirements, Cn_max is the preset maximum number of process parameters, Rn_max is the preset maximum number of special reagent types, and En_max is the preset maximum number of equipment function requirements. Calculate the priority score Sp, where Sp = (d × Wd + e × We + f × Wc) × K, where d, e, and f are the priority dimension weights, corresponding to the influence weights of service level, order urgency, and demand complexity, respectively, and d + e + f = 1; K is the core demand adaptation correction item, determined based on the individual adaptation score Sfi of the core demand tag. For example, if the Sfi of all core demand tags is ≥ 0.6, then K = 1.1; if there is one core demand tag with Sfi < 0.6, then K = 0.9; if there are two or more core demand tags with Sfi < 0.6, then K = 0.7; obtain the value range of the priority score Sp, and divide it equally according to the number M of preset sorting priorities to obtain the value interval corresponding to each preset sorting priority; compare the calculated priority score Sp with the value interval corresponding to each preset sorting priority, output the sorting priority level, and store it in the system along with the unique clothing number.

[0031] In this embodiment, it is assumed that the priority score Sp ranges from [Sp_min, Sp_max], where Sp_min is the minimum Sp value of all orders in the system in the past 3 months, and is 0 when < 0; Sp_max is the maximum Sp value of all orders in the system in the past 3 months; the number of sorting priorities is preset to M, where M is a positive integer ≥ 2, and is preset and configured by the system according to the number of sorting equipment channels and the scale of workstations. Calculate the value range corresponding to each priority: the interval length ΔS = (Sp_max - Sp_min) / M. Divide the intervals in order from "highest to lowest". The value range of the first priority (highest) is (Sp_max - ΔS, Sp_max], the value range of the second priority is (Sp_max - 2 × ΔS, Sp_max - ΔS], and so on. The value range of the Mth priority (lowest) is [Sp_min, Sp_max - (M-1) × ΔS].

[0032] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0033] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A clothing sorting and management method based on artificial intelligence, characterized in that: The method includes the following steps: S1. Obtain the order information corresponding to the clothes to be sorted. The order information includes customer type identifier, order service demand coefficient, preset delivery time limit and service demand terms. Based on the customer type identifier, associate the preset customer type and obtain the customer demand tag that is preset to be bound to the corresponding customer type. The demand tag includes functional constraint threshold, care level standard, sorting isolation requirements and packaging specifications. S2. Collect the attribute information of the clothes to be sorted. The attribute information includes the type of clothing material, the classification of uses, and historical washing records. The historical washing records include the previous adapted process parameters, damage feedback data, and compatibility scores. S3. Cross-analyze customer demand tags and clothing attribute information, calculate the compatibility score between each functional constraint threshold in the demand tags and clothing attributes, and generate corresponding clothing washing adaptation schemes based on the compatibility score. The adaptation schemes include process parameter ranges, special reagent types and equipment functional requirements. S4. Determine the service level weight based on the order service demand coefficient in the order information, and calculate the order urgency coefficient; at the same time, calculate the demand complexity coefficient of the washing adaptation solution, and obtain the corresponding priority score, and divide the sorting priority according to the priority score; S5. Based on sorting priority and washing adaptation scheme, control the sorting equipment to transport the clothes to be sorted to the dedicated sorting channel and target workstation that match the functional requirements of the equipment, and simultaneously record the correlation data of clothing attributes, demand tags, priority scores and sorting paths, and feed back the sorting completion status and channel matching results to the system in real time.

2. The clothing sorting and management method based on artificial intelligence according to claim 1, characterized in that: S1 includes the following: The system interface obtains the order information corresponding to the clothes to be sorted, and extracts the customer type identifier, order service demand coefficient, preset delivery time limit and service demand terms from the order information; the customer type identifier is a preset unique code, the order service demand coefficient is a quantitative value of 1-5 levels, and the service demand terms include quantitative indicators related to functional constraint thresholds. A mapping database is constructed between customer type identifiers and preset customer types. This database stores a one-to-one correspondence between preset codes and preset customer types. The extracted customer type identifiers are precisely matched with the preset codes in the mapping database. If a match is successful, the corresponding preset customer type is directly determined. If a match fails, core functional keywords from the service requirement terms are extracted using a preset keyword dictionary, which contains typical functional requirement keywords corresponding to each preset customer type. The extracted core functional keywords are then matched with the typical requirement keywords of each preset customer type based on similarity. The similarity calculation formula is: S = (M / N) × 100%, where M is the number of successfully matched core keywords and N is the total number of core keywords in the service requirement terms. If S ≥ 80%, the corresponding preset customer type is determined. If S < 80%, a customer type matching failure message is generated and fed back to the system in real time. The order sorting process is paused until manual verification is performed before re-executing the matching step. The service level base weight Wb is calculated based on the order service demand coefficient, and Wb = L / Lmax, where L is the order service demand coefficient, L∈[1,5], and Lmax is the maximum service level parameter and Lmax=5; a predefined customer type and customer demand tag association database is constructed. Each predefined customer type in the association database corresponds to a set of weighted demand tags. The sum of the weights of each demand tag is 1. The weight of core demand tags is ≥60%, and the weight of non-core demand tags is ≤40%; according to the determined predefined customer type, the corresponding demand tag group is retrieved from the association database as the customer demand tag.

3. The clothing sorting and management method based on artificial intelligence according to claim 2, characterized in that: S2 includes the following: By collecting basic attribute information and garment label images of the clothes to be sorted, and combining this with the material descriptions provided by customers in the order information, the material type and usage classification of the clothes are jointly determined. The material type is determined through a triple verification process: extracting visual features of the garment's fiber texture and surface gloss, identifying the preset material identifier in the label image, verifying consistency with the order's material description, and encoding it according to preset classification rules from M1 to Mn, where n is the total number of material types. Usage classification is determined by identifying structural features of the garment's collar style, pocket design, and size specifications, matching them with keywords used in the scene description in the order, and encoding it according to preset classification rules from U1 to Um, where m is the total number of usage classifications. The classification by application is achieved through the following steps: The system performs contour segmentation, edge detection, and region labeling on clothing to extract structural features such as neckline style, number of pockets, size specifications, and overall contour shape. These extracted features are then converted into structured feature items for neckline type, pocket presence / absence, size range, and contour shape. The system also segments the scenario description text used in orders, removing stop words and extracting usage keywords such as bedding, tops, trousers, hotel linens, and business formal wear. A mapping rule base between structural features and usage keywords is established. The extracted structured feature items are matched against this rule base item by item, and the number of successfully matched feature items is counted. The feature matching degree is calculated, and when the feature matching degree exceeds a preset threshold, the usage classification is completed. The unique garment IDs are extracted from the obtained order information, and a database linking these unique garment IDs with historical washing records is constructed. This database stores the previous adaptation process parameters, damage feedback data, and single-time compatibility scores corresponding to each garment ID. The extracted garment IDs are then precisely matched with the database. If a match is successful, all corresponding historical washing records are retrieved. If a match fails, a multi-dimensional feature combination of the garment is extracted using a visual acquisition device and compared with the feature combinations of historical garments in the database. The similarity calculation formula is Sc = (P / Q) × 100%, where P is the number of successfully matched feature items and Q is the total number of feature items in the multi-dimensional feature combination. When Sc ≥ 95%, the corresponding historical washing record is retrieved. The overall garment compatibility score Pa is calculated based on retrieved historical washing records. The formula is: Pa = α × A + β × (1 - B), where α is the process compatibility weight, β is the damage impact weight, and α + β = 1, with specific values ​​determined based on garment material type and customer requirement labels; A is the average compatibility between past compatible processes and garment materials, and A = ∑Ai / k, where k is the number of historical washes, Ai is the compatibility score for the i-th compatible process, Ai = γ × C + δ × D, where γ is the process parameter matching weight, δ is the washing effect weight, and... γ+δ=1, the specific value is determined according to the clothing use classification and the functional constraint threshold in the customer demand label; C is the matching degree of process parameters, and C=(number of parameters in the matching process that meet the requirements of clothing material / total number of process parameters)×100%, C∈[0,1]; D is the washing effect compliance rate, D=(total area of ​​stains on clothing before washing-total area of ​​stains on clothing after washing) / total area of ​​stains on clothing before washing, D∈[0,1]; B is the average proportion of damaged area in historical washing, B=∑Bi / k, Bi is the proportion of damaged area in the i-th washing. By integrating material type coding, usage classification coding, compatibility comprehensive score Pa, previous compatibility process parameters, and average damage area percentage B, a complete dataset of clothing attribute information is formed.

4. The clothing sorting and management method based on artificial intelligence according to claim 3, characterized in that: S3 includes the following: The customer demand tags and clothing attribute information are preprocessed in a structured manner to extract the determined customer demand tag groups and clarify the target weight ω of each demand tag. The target weight comes from the preset weight of the S1 associated database. Extract the resulting clothing attribute information dataset, filter the core attributes associated with the demand tags, specifically including: material type code M, usage classification code U, overall suitability score Pa, average damaged area percentage B, and previous suitability process parameter set; establish a cross-analysis dimension mapping table to correspond the functional constraint thresholds in the demand tags with the core attributes of the clothing, forming a mapping relationship between functional constraint types and clothing attribute dimensions. For each functional constraint threshold, combined with its corresponding clothing attribute dimension, calculate the individual fit score Sf for that clothing attribute. The specific formula is: Sf=ω×θ×[φ×F(M,U)+σ×G(Pa,B)], where θ is the functional constraint type coefficient, φ is the material and usage fit coefficient, determined according to the combination relationship between material type M and usage category U; F(M,U) is the basic fit value for material and usage, F(M,U) = (the baseline score of material M fitting this constraint) +The baseline score for the application U to adapt to this constraint) / 2, the baseline score is preset to 0-1 points; σ is the historical data correction coefficient, which is dynamically adjusted according to the comprehensive adaptability score Pa and the average damage area ratio B: σ=1+0.05×(Pa-0.5)-0.1×B, if σ<0.9 then take 0.9, σ>1.1 then take 1.1; G(Pa,B) is the historical data adaptation value, G(Pa,B)=Pa×(1-B), G(Pa,B)∈[0,1]; Calculate the overall fit score St and determine the fit level. The formula for calculating the overall fit score St is: St=∑(Sfi×ωi) / ∑ωi, where Sfi is the fit score of the i-th functional constraint, ωi is the label weight corresponding to the i-th functional constraint, and St∈[0,1]. When St≥0.8, it is marked as Level 1 fit; when 0.6≤St<0.8, it is marked as Level 2 fit; when St<0.6, it is marked as Level 3 fit; if the Sfi corresponding to the core requirement label is <0.5, the fit level is directly reduced to Level 3. Based on the adaptation level and garment attributes, a garment washing adaptation plan is generated. For Level 1 adaptation, the previously retrieved adaptation process parameters are used as the benchmark, with a fluctuation range of ≤±5%. For Level 2 adaptation, the standard process parameters corresponding to material M are used as the benchmark, combined with adjustments based on functional constraint thresholds, with a fluctuation range of ≤±10%, and the constraint thresholds of the core requirement label must be met. For Level 3 adaptation, the functional constraint thresholds of the core requirement label are prioritized, with a fluctuation range of ≤±15% for process parameters, while reducing the washing intensity. If the average damaged area ratio B ≥ 0.1, the process parameter range is further widened by 5% based on the corresponding adaptation level. According to the material M, functional constraint thresholds, and adaptation level, Level 1 adaptation uses previously adapted reagents, Level 2 adaptation uses general-purpose special reagents, and Level 3 adaptation uses mild special reagents, while avoiding incompatible reagents associated with historical damage feedback data. Based on the process parameter range, material M, and sorting and isolation requirements, the functions that the equipment must possess are clearly defined. Integrate process parameter ranges, special reagent types, and equipment functional requirements to form a structured adaptation solution document, associate it with the unique garment number and total compatibility score St, and store it in the system database.

5. The clothing sorting and management method based on artificial intelligence according to claim 4, characterized in that: S4 includes the following: Extract the order service demand coefficient L and the preset delivery time limit T0. Extract the core parameters in the generated washing adaptation plan. At the same time, obtain the current system time T1 and the order creation time Ts. Calculate the order's elapsed time Tu, and Tu = T1 - Ts. The service level base weight Wb is calculated based on the customer type correction coefficient λ, and the service level weight Wd is calculated by introducing the customer type correction coefficient λ. Wd = Wb × λ, where the customer type correction coefficient is set according to the preset customer type. The urgency level is quantified based on the difference between the preset delivery time limit T0 and the order's elapsed time Tu. The order urgency coefficient We is calculated, and We = 1 / [(T0-Tu)+1]×100%; If T0 - Tu ≤ 0, then We = 100%; The complexity coefficient Wc of sorting requirements is calculated based on the core parameters of the washing adaptation scheme. The calculation formula is: Wc=(a×Cn+b×Rn+c×En) / (Cn_max+Rn_max+En_max), where a, b, and c are weight coefficients, corresponding to the influence weights of process parameters, reagent types, and equipment functions, respectively, and a+b+c=1; Cn is the number of process parameters, Rn is the number of special reagent types, En is the number of equipment function requirements, Cn_max is the preset maximum number of process parameters, Rn_max is the preset maximum number of special reagent types, and En_max is the preset maximum number of equipment function requirements. Calculate the priority score Sp, where Sp = (d × Wd + e × We + f × Wc) × K, where d, e, and f are the priority dimension weights, corresponding to the influence weights of service level, order urgency, and demand complexity, respectively, and d + e + f = 1; K is the core demand adaptation correction item, determined based on the individual adaptation score Sfi of the core demand tag; obtain the value range of the priority score Sp, and divide it equally according to the number M of preset sorting priorities to obtain the value interval corresponding to each preset sorting priority; compare the calculated priority score Sp with the value interval corresponding to each preset sorting priority, output the sorting priority level, and store it in the system along with the unique clothing number.

6. An artificial intelligence-based clothing sorting and management system, applied to the artificial intelligence-based clothing sorting and management method according to any one of claims 1-5, characterized in that: The system includes: an order information parsing and demand tag matching module, a clothing attribute collection and historical data retrieval module, an adaptation scheme generation module, a sorting priority calculation module, and a sorting execution and data feedback module. The order information parsing and demand tag matching module is used to obtain the order information of the clothes to be sorted, parse and extract key parameters and associate them with preset customer types, and retrieve the corresponding customer demand tags. The clothing attribute acquisition and historical data retrieval module is used to collect basic attribute information and washing label images of clothing, determine the material type and usage classification, retrieve the corresponding historical washing records, and integrate them to form an attribute information dataset. The adaptation scheme generation module is used to cross-analyze customer demand tags and clothing attribute information, calculate the adaptation score, and generate a washing adaptation scheme that includes process parameter range, special reagent type, and equipment functional requirements. The sorting priority calculation module is used to calculate the service level weight, order urgency coefficient, and demand complexity coefficient based on order information and washing adaptation plan, to obtain a priority score and classify sorting priorities; The sorting execution and data feedback module is used to control the operation of the sorting equipment based on sorting priority and adaptation scheme, record related data, and provide real-time feedback on sorting status and channel matching results.

7. The clothing sorting and management system based on artificial intelligence according to claim 6, characterized in that: The order information parsing and demand tag matching module includes an order information acquisition unit and a demand tag matching unit; The order information acquisition unit obtains the order information corresponding to the clothes to be sorted through the system interface, and parses and extracts the customer type identifier, order service demand coefficient, preset delivery time limit and service demand terms to complete the validity verification of the order information. The demand tag matching unit matches customer type identifiers with preset customer types, determines customer types through precise matching in a mapping database or keyword similarity matching, and retrieves a weighted demand tag group that is pre-bound to that customer type.

8. The clothing sorting and management system based on artificial intelligence according to claim 6, characterized in that: The clothing attribute acquisition and historical data retrieval module includes a clothing attribute acquisition unit and a historical data retrieval unit; The clothing attribute acquisition unit collects basic attribute information and washing label images of clothing through a visual acquisition device. Combined with the material description filled in by the customer in the order, it determines the clothing material type through triple verification and classifies the usage by matching structural features and keywords. The historical data retrieval unit extracts the unique number of clothing from the order information, retrieves historical washing records through precise matching or multi-dimensional feature combination similarity matching, calculates the comprehensive clothing suitability score, and integrates them to form a complete clothing attribute information dataset.

9. The clothing sorting and management system based on artificial intelligence according to claim 6, characterized in that: The adaptation scheme generation module includes a cross-analysis unit and an adaptation scheme output unit; The cross-analysis unit performs structured preprocessing on customer demand tags and clothing attribute information, establishes a cross-analysis dimension mapping relationship, calculates the individual and total fit scores of each functional constraint threshold and clothing attribute, and determines the fit level. The fit scheme output unit determines the range of process parameters, the type of special reagents, and the equipment functional requirements based on the fit level, clothing material type, and historical damage status, avoids prohibited reagents, integrates and forms a structured clothing washing fit scheme, and stores it.

10. The clothing sorting and management system based on artificial intelligence according to claim 6, characterized in that: The sorting priority calculation module includes a priority parameter calculation unit and a priority division unit, and the sorting execution and data feedback module includes a sorting control unit and a data recording feedback unit. The priority parameter calculation unit extracts the core parameters from the order and adaptation plan, calculates the service level weight, order urgency coefficient and requirement complexity coefficient, and obtains the priority score by combining the core requirement adaptation correction items. The priority division unit obtains the value range of the priority score, divides the interval equally according to the preset sorting priority quantity, and determines the sorting priority level by score comparison; The sorting control unit controls the sorting equipment to transport clothes to the matching dedicated sorting channel and target workstation based on sorting priority and washing adaptation scheme; the data recording and feedback unit synchronously records the correlation data of clothing attributes, demand tags, priority scores and sorting paths, and feeds back the sorting completion status and channel matching results to the system in real time.

Images