Energy distribution method, equipment and system for article processing

Abstract

The invention relates to the field of computer technology application, in particular to an energy distribution method, equipment and system for article processing, and the method is realized based on an article processing record data information table RF. The RF divides articles into different clusters and sub-clusters according to visual perception attributes, and records key information such as a single-batch maximum processing weight threshold, an energy consumption weight correlation expression and a cleanliness weight correlation expression of each article sub-cluster in different processing modes. According to the method, the processing information for the current to-be-processed article is generated by acquiring the data record table of the current to-be-processed article and combining with the information in the RF, so that accurate distribution of energy is realized. According to the method, the energy utilization efficiency can be improved, the stability of the processing effect is guaranteed, the intelligence and automation level of the processing flow is improved, and the method has good adaptability and expansibility and is suitable for various article processing scenes.

Description

Technical Field

[0001] The present invention relates to the field of computer technology applications, and in particular to an energy distribution method, device and system for article processing. Background Art

[0002] With the rapid development of industrial automation and intelligent manufacturing, item handling systems (such as laundry systems) are becoming increasingly common in commercial and household laundry applications. Laundry systems, as typical high-energy-consuming item handling scenarios, encompass multiple steps, including clothing sorting, pre-treatment, washing, drying, ironing, and disinfection. The energy consumption of each step (e.g., electric motors, heating devices, and steam systems for heat supply) must be coordinated through computer system scheduling. However, the energy allocation logic of current laundry systems still relies on empirically defined, fixed procedures and lacks dynamic optimization capabilities based on data processing. This results in low energy efficiency. Specifically, energy allocation parameters (such as wash water temperature, motor speed, and drying time) in existing systems are often preset based on manual experience, rather than precisely controlled through sensor data and algorithmic analysis. For example, the fixed parameter "heating cotton clothes at 60°C" in a wash program fails to account for details such as the actual level of soiling (light sweat stains do not require high temperatures) and the thickness of the clothes (the difference in thermal energy requirements between thin T-shirts and thick coats). This leads to a disconnect between the energy instructions executed by the computer system and actual demand. Under this empirical logic, the ineffective energy consumption of a single wash can reach 15%-25%, which is essentially due to the lack of real-time processing and parameter mapping capabilities of clothing characteristic data. Summary of the Invention

[0003] In view of the above technical problems, the technical solution adopted by the present invention is: According to a first aspect of the present invention, a method for allocating energy for article processing is provided. The method is implemented based on a data processing system. The data processing system includes a storage module. The storage module stores an article processing record data information table RF, where RF = {RF1, RF2, ..., RF u ,……,RF q}, RF u is the record information of the u-th item cluster, where u ranges from 1 to q, and q is the number of item clusters. Item clusters are divided based on the visual perception attributes of items; RF u ={RF u1 ,……,RF uv ,……,RF uf（u）}, RF uv For RF u The record information of the vth item sub-cluster in , where v ranges from 1 to f(u), and f(u) is RF u The number of item subclusters in RF uv ={U uv, RF uv 1 ,……,RF uv r ,……,RF uzv z（u，v）}, U uv is the unique identifier of the vth item sub-cluster, RF uv r is the record information of the v-th item sub-cluster under the r-th processing method, where r ranges from 1 to z(u, v), and z(u, v) is the number of processing methods for the v-th item sub-cluster; RF uv r =(UT uv r , WT uv r , YE uv r , YC uv r ), U uv r is the unique identifier of the rth processing method, WT uv r YE is the maximum processing weight threshold of a single batch of the vth item sub-cluster under the rth processing method, uv r is the energy consumption-weight correlation expression of the vth item sub-cluster under the rth treatment mode, YC uv r is the cleanliness weight association expression of the vth item subcluster under the rth processing mode; the method comprises the following steps: S100. The data processing system obtains the digitized visual feature data and the numerical processing weight data of the current item to be processed through the data acquisition interface to generate a structured data record table, which is stored in the storage module; each row of data in the data record table corresponds to an item subcluster, and each row of data includes: a unique identifier of the corresponding item subcluster and the processing weight of the item to be processed in the corresponding item subcluster.

[0004] S200, the data processing system calls the item processing record data information table and the data record table in the storage module, generates processing information for the current item to be processed, and outputs the processing information, wherein the processing information includes processing mode and processing batch information.

[0005] According to a second aspect of the present invention, an electronic device is provided, comprising a processor and a memory; the processor is configured to execute the steps of the method according to the first aspect of the present invention by calling a program or instruction stored in the memory.

[0006] According to a third aspect of the present invention, there is provided an energy distribution system for article processing, the system comprising: Storage module, the storage module stores the article processing record data information table RF; RF={RF1, RF2, ..., RF u ,……,RF q}, RF u is the record information of the u-th item cluster, where u ranges from 1 to q, and q is the number of item clusters. Item clusters are divided based on the visual perception attributes of items; RF u ={RF u1 ,……,RF uv ,……,RF uf（u）}, RF uv For RF u The record information of the vth item sub-cluster in , where v ranges from 1 to f(u), and f(u) is RF u The number of item subclusters in RF uv ={U uv , RF uv 1 ,……,RF uv r ,……,RF uzv z（u，v）}, U uv is the unique identifier of the vth item sub-cluster, RF uv r is the record information of the v-th item sub-cluster under the r-th processing method, where r ranges from 1 to z(u, v), and z(u, v) is the number of processing methods for the v-th item sub-cluster; RF uv r =(UT uv r , WT uv r , YE uv r , YC uv r ), U uv r is the unique identifier of the rth processing method, WT uv r YE is the maximum processing weight threshold of a single batch of the vth item sub-cluster under the rth processing method, uv r is the energy consumption-weight correlation expression of the vth item sub-cluster under the rth treatment mode, YC uv r is the cleanliness weight association expression of the vth item subcluster under the rth treatment method.

[0007] The data acquisition module is used to obtain the digitized visual feature data and the numerical processing weight data of the current items to be processed through the data acquisition interface to generate a structured data record table for storage in the storage module; wherein each row of data in the data record table corresponds to an item subcluster, and each row of data includes: a unique identifier of the corresponding item subcluster and the processing weight of the items to be processed in the corresponding item subcluster.

[0008] The processing information generation module is used to call the item processing record data information table and the data record table in the storage module, generate processing information for the current item to be processed, and output the processing information, wherein the processing information includes processing mode and processing batch information.

[0009] An energy allocation method for item processing provided by an embodiment of the present invention includes: obtaining a data record table currently to be processed, the data record table being generated based on the visual features of the items currently to be processed; generating processing information for the items currently to be processed based on RF and the data record table, and outputting the processing information, the processing information including processing method and processing batch information. This method has at least the following technical effects: (1) Achieve precise energy allocation: Relying on the item processing record data information table RF, the items are subdivided into clusters and subclusters. The subdivision is based on the visual perception attributes of the items, which can more accurately match the characteristics of different item subclusters. Combining the maximum processing weight threshold, energy consumption weight correlation expression, and cleanliness weight correlation expression of each subcluster under different processing methods, the optimal energy allocation plan can be calculated for the specific weight of the current item to be processed, avoiding energy waste and improving energy utilization efficiency.

[0010] (2) Ensure the stability of treatment effects: The historical data stored in the RF can provide a reliable reference for current treatment. The cleanliness weight correlation expression can ensure that when allocating energy, the cleanliness of the items will not be reduced due to excessive pursuit of energy saving. The treated items will always maintain a high level of cleanliness, ensuring the stability of the treatment effect.

[0011] (3) Improve the intelligence and automation level of the processing process: Based on the data record table and RF information table, processing information can be automatically generated to reduce manual intervention. There is no need for manual experience to judge the energy allocation amount, which reduces human errors, makes the item processing process more intelligent and automated, and improves processing efficiency.

[0012] (4) Enhanced adaptability and scalability of the method: The division of item clusters and subclusters and the structural design of the RF information table enable the method to adapt to the processing requirements of different types of items. When new item types or processing methods emerge, the RF information table only needs to be updated and expanded accordingly to incorporate the new content into the energy distribution system, which has strong adaptability and scalability.

[0013] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0015] Figure 1 This is a flow chart of an energy distribution method for article processing provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention pertains. The terms used herein in the specification of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0018] It should be noted that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the steps as sequential processes, many of the steps can be performed in parallel, concurrently, or simultaneously. In addition, the order of the steps can be rearranged. A process can be terminated when its operation is completed, but can also have additional steps not included in the accompanying drawings. A process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0019] An embodiment of the present invention provides an energy allocation method for article processing, wherein the method is implemented based on a data processing system. The data processing system includes a storage module, wherein the storage module stores an article processing record data information table RF.

[0020] Where RF={RF1, RF2, ..., RF u ,……,RF q}, RF u is the record information of the u-th item cluster, where u ranges from 1 to q, and q is the number of item clusters. Item clusters are divided based on the visual perception attributes of items; RF u ={RF u1 ,……,RF uv ,……,RF uf（u）}, RF uv For RF u The record information of the vth item sub-cluster in , where v ranges from 1 to f(u), and f(u) is RF u The number of item subclusters in RF uv ={U uv , RF uv 1 ,……,RF uv r ,……,RF uzv z（u，v）}, U uv is the unique identifier of the vth item sub-cluster, RF uv r is the record information of the v-th item sub-cluster under the r-th processing method, where r ranges from 1 to z(u, v), and z(u, v) is the number of processing methods for the v-th item sub-cluster; RF uv r =(UT uv r , WT uv r , YE uv r , YC uv r ), U uv r is the unique identifier of the rth processing method, WT uv r YE is the maximum processing weight threshold of a single batch of the vth item sub-cluster under the rth processing method, uv r is the energy consumption-weight correlation expression of the vth item sub-cluster under the rth processing mode, which can be stored in the storage module in the form of function pointer. uv rThe cleanliness weight association expression of the v-th item sub-cluster under the r-th processing method can be stored in the form of an executable code fragment.

[0021] In this embodiment of the present invention, the visually perceived attribute may be color. Object clusters are generated by applying a K-means clustering algorithm to the digital features of the visually perceived attributes of the objects (e.g., color pixel value matrices and texture feature vectors). The clustering process iteratively optimizes cluster centers to minimize the variance of the data within the cluster.

[0022] In an embodiment of the present invention, the item type includes dimensions, material, and a soiling parameter value. The unique identifier of an item subcluster can be a code value for the item subcluster. This code value is unique and identifiable and can be in the form of letters, numbers, or a combination thereof, used to uniquely distinguish different item subclusters. For example, "C01-S03" can represent the third item subcluster under the first item cluster. The unique identifier of a treatment method is the code value for the treatment method. This code value is also unique and can be compiled based on characteristics such as the type and parameters of the treatment method to quickly identify the corresponding treatment method. For example, "P02-H05" can represent the fifth sub-treatment method within the second category of basic treatment methods. The above code values ​​should remain globally unique in the item treatment record data information table RF to ensure that they can be accurately associated with the corresponding item subcluster and treatment method during data storage, retrieval, and processing information generation to avoid confusion.

[0023] In embodiments of the present invention, the soiling parameter values ​​for dimensional specifications can be determined using manual labeling or visual inspection. Manual labeling involves an operator assigning a grading system (e.g., "light," "moderate," or "heavy") to the soiling level of an item based on experience, then assigning a fixed parameter value (e.g., 20, 50, or 80) to each level. To reduce subjective error, a standard reference chart (e.g., sample images of varying soiling levels) can be developed to ensure more consistent labeling. The visual inspection method utilizes an image acquisition device (e.g., a webcam or industrial camera) to capture an image of the item's surface, analyze the characteristics of the soiled areas within the image (e.g., area, color depth, contrast, etc.), and convert this image into a soiling parameter value. Quantification involves setting a "cleanliness benchmark" (e.g., the RGB mean value for pure white clothing), calculating the deviation between the actual image and the benchmark (e.g., color deviation percentage, grayscale value difference), and calculating the percentage of soiled area on the item's surface area (e.g., "5%" corresponds to low soiling, "30%" corresponds to high soiling). This is then mapped to a numerical value on a scale of 0-100 (0 being completely clean and 100 being severely soiled).

[0024] In the embodiment of the present invention, the item sub-clustering is implemented by the data processing system through a hierarchical clustering algorithm based on the digital coding of the item type characteristic parameters, and the characteristic similarity of the data in the sub-cluster is ≥90%.

[0025] In the embodiment of the present invention, the record information of the processing method is stored in a standardized data structure.

[0026] In an exemplary embodiment of the present invention, the article processing process is to perform a processing operation on the article by a processing device, the article is specifically clothes that need to be washed, and the processing device is a washing machine.

[0027] Furthermore, in an embodiment of the present invention, the processing mode for each item sub-cluster is a processing mode that can process the item sub-cluster. In the case where the items are clothes, the processing mode may be a washing mode. The unit of processing weight is a preset unit, such as kg.

[0028] Furthermore, in the embodiment of the present invention, YE uv r and YC uv r This can be obtained based on the historical data set for the vth item subcluster under the rth processing method. Each historical data point in the historical data set includes the processing weight, energy consumption parameter value, and cleanliness parameter value. The historical data set can be obtained based on the operation logs of item processing equipment (such as washing machines and dishwashers).

[0029] In the embodiment of the present invention, the energy consumption parameter value E corresponding to the dth historical data d The following conditions are met: d =(Ea d -Eb d ) / Eb d ×100, Ea d is the actual energy consumption value corresponding to the processing weight in the dth historical data, Eb d is the benchmark energy consumption value corresponding to the processing weight in the d-th historical data, d ranges from 1 to Q, and Q is the number of historical data.

[0030] In the embodiment of the present invention, Ea d It can be equal to the product of the actual water consumption corresponding to the treatment weight in the dth historical data and the water conversion coefficient plus the product of the actual electricity consumption corresponding to the treatment weight in the dth historical data and the electricity conversion coefficient. The water conversion coefficient can be approximately equal to 0.000086, and the electricity conversion coefficient can be approximately equal to 0.1229. d It can be an empirical value, which can be obtained by constructing a benchmark mapping relationship of "processing weight-benchmark energy consumption value" based on historical operating data through data processing and algorithm modeling.

[0031] In an embodiment of the present invention, the cleanliness parameter value can be obtained based on a cleanliness detector or a sensory score. The maximum cleanliness parameter value can be set to 10 points. The higher the cleanliness, the larger the corresponding cleanliness parameter value.

[0032] Specific, YEuv r It can be obtained by the following method: pre-process the historical data set of the v-th item sub-cluster under the r-th processing method, and remove the data with abnormal processing weight or energy consumption parameter values ​​(the criterion for abnormal data is that the deviation from the mean of the data set exceeds 3 times the standard deviation); use the processing weight in the pre-processed data set as the independent variable and the energy consumption parameter value as the dependent variable, and use the least squares method to perform linear fitting to obtain the linear regression equation YE uv r =k1×N+b1, where k1 is the fitting slope, b1 is the intercept, N is the treatment weight, and the goodness of fit R² must be ≥0.9.

[0033] YC uv r It is obtained by the following method: pre-processing the historical data set of the v-th item sub-cluster under the r-th processing method, eliminating the data with abnormal processing weight or cleanliness parameter values ​​(the abnormal data judgment criteria are the same as above); using the processing weight in the pre-processed data set as the independent variable and the cleanliness parameter value as the dependent variable, the least squares method is used for linear fitting to obtain the linear regression equation YC uv r =k²×N+b², where k² is the fitting slope and b² is the intercept, and the goodness of fit R² must be ≥0.9. It should be noted that in actual applications, if the predicted cleanliness parameter value in the fitting result exceeds the maximum cleanliness parameter value Cmax, the predicted cleanliness parameter value is taken as Cmax.

[0034] In an embodiment of the present invention, the maximum processing weight threshold of a certain processing method refers to the maximum weight limit of items that can be processed in a single batch under this processing method, that is, the upper limit of the weight that cannot be exceeded in a single processing under the premise of ensuring the preset processing effect (such as meeting the cleanliness requirements, energy consumption within a reasonable range, etc.). Taking the item processing scenario as an example, if the maximum processing weight threshold of a certain processing method is 10kg, it means that when using this method to process items in a single batch, the processing weight is at most 10kg. When the weight of the items to be processed does not exceed 10kg, the processing can be completed in a single batch and the processing effect can be guaranteed; if it exceeds 10kg, single-batch processing may result in substandard cleanliness and abnormally increased energy consumption. In this case, it is necessary to process in multiple batches, and the processing weight of each batch must not exceed the threshold.

[0035] Furthermore, WT uv r Obtain it in the following ways: S1, obtain the historical data set of the vth item sub-cluster under the rth processing method, each historical data in the historical data set includes the processing weight, energy consumption parameter value and cleanliness parameter value.

[0036] S2, deleting the historical data whose cleanliness parameter value is less than the set cleanliness parameter value threshold in the historical data set, and obtaining the deleted historical data set as the target data set.

[0037] S3. For each data point in the target dataset, obtain the corresponding fusion value. The fusion value satisfies the following condition: SM = We × E + Wc × (Cmax - C), where SM is the fusion value corresponding to the data point, used to quantify the comprehensive performance of energy consumption and cleanliness. E is the energy consumption parameter value corresponding to the data point. C is the cleanliness parameter value corresponding to the data point. We is the energy consumption weight, Wc is the cleanliness weight, and We + Wc = 1.

[0038] In an embodiment of the present invention, the values ​​of We and Wc can be dynamically adjusted based on actual needs: if energy consumption needs to be reduced in priority in the processing scenario (such as in energy-saving mode), We>Wc is set (for example, We=0.6, Wc=0.4); if cleanliness needs to be guaranteed in priority in the processing scenario (such as in deep cleaning mode), We<Wc is set (for example, We=0.3, Wc=0.7); if the importance of energy consumption and cleanliness needs to be balanced (such as in standard mode), We=Wc is set (for example, We=Wc=0.5).

[0039] S4, using the processed weight in the target data set as the independent variable and the fused value in the target data set as the dependent variable, obtains the data fitting curve.

[0040] In embodiments of the present invention, a polynomial or nonlinear fitting method can be used to construct the curve. The fitting function type must be able to reflect the trend relationship between the processed weight and the fusion value. In one exemplary embodiment, a third-order polynomial function can be used, expressed as SM = a × N³ + b × N² + c × N + d, where a, b, c, and d are fitting coefficients. The fitting process uses the least squares method to solve the coefficients, and the goodness of fit R² must be ≥ 0.9 to ensure that the curve is representative of the data trend.

[0041] It should be noted that if the amount of data in the target data set is less than the preset number, such as less than 30 data points, interpolation can be used to supplement the data before fitting to ensure the stability of the curve.

[0042] S5, obtaining the treatment weight corresponding to the slope mutation point of the data fitting curve as WT uv r .

[0043] In the embodiment of the present invention, the slope mutation point satisfies the following conditions: In the neighborhood to the left of the treatment weight corresponding to the slope mutation point, the absolute value of the slope of the data fitting curve is less than or equal to K1, and the slope sign is negative or close to zero. This means that the fusion value SM shows a significant upward trend with increasing treatment weight. K1 is the preset flat slope threshold.

[0044] In the neighborhood to the right of the treatment weight corresponding to the slope mutation point, the absolute value of the slope of the data fitting curve is greater than or equal to K2, and the slope sign is positive, indicating that the fusion value SM shows a significant upward trend with increasing treatment weight. K2 is the preset steep slope threshold, and K2>K1.

[0045] The slope mutation point is the turning point where the first-order derivative of the curve suddenly changes from less than or equal to K1 to greater than or equal to K2, and the second-order derivative at the turning point is greater than or equal to M, where M is a preset second-order derivative threshold.

[0046] The values ​​of K1, K2, and M need to be calibrated through experimental data based on the specific processing scenario (such as item type and processing technology). For example, for textile items, K1=0.1, K2=0.5, and M=0.3 can be set.

[0047] Those skilled in the art should understand that the method for obtaining the slope mutation point of the data fitting curve may fall within the scope of the existing technology.

[0048] Furthermore, an embodiment of the present invention provides an energy distribution method for article processing, which may include: Figure 1 The following steps are shown: At step S100, the data processing system acquires digitized visual feature data and numerical data on the processed weight of the current item to be processed through the data acquisition interface to generate a structured data record table for storage in the storage module. In this embodiment of the present invention, visual features may include visual perception attributes, material, dimensions, and contamination level. The specific logic for generating the data record table is as follows: Visual features of the items to be processed are extracted through image recognition, sensor detection, and other technical means (e.g., determining material through spectral analysis, obtaining dimensions through image size measurement, and determining contamination level through grayscale analysis). The extracted features are then compared and matched with a pre-set item sub-cluster feature library to determine the item sub-cluster to which each item to be processed belongs. The relevant information is then categorized and recorded by item sub-cluster.

[0049] In addition, the data record table may additionally include auxiliary information such as record generation time, feature recognition confidence (such as the probability value of successful matching), data verification code, etc. The feature recognition confidence is used to evaluate the reliability of the sub-cluster matching results. When the confidence is lower than the preset threshold (such as 85%), manual review is required before generating the formal data record table to ensure the accuracy of the data basis.

[0050] In the embodiment of the present invention, each row of data in the data recording table corresponds to an item sub-cluster, and each row of data includes: a unique identifier of the corresponding item sub-cluster and a processing weight of the items to be processed in the corresponding item sub-cluster.

[0051] S200, the data processing system calls the item processing record data information table and the data record table in the storage module, generates processing information for the current item to be processed, and outputs the processing information, wherein the processing information includes processing mode and processing batch information.

[0052] Those skilled in the art should understand that the processing information for the current item to be processed includes the processing information of the item to be processed corresponding to each row of data in the data recording table.

[0053] In an embodiment of the present invention, the data processing system is a functional unit for performing electronic digital data processing, which may include a data acquisition module, a data operation module, a storage module, and an output module. Each module realizes data interaction through an internal data link, which is specifically defined as follows: Data acquisition module: responsible for receiving visual feature digitized data (such as pixel value array, feature vector) and processing weight digitized data through a data interface, converting the analog signal into a digital signal and transmitting it to the data operation submodule. Data operation module: equipped with a data processing algorithm, it can perform cluster analysis, association expression calculation, comparison and screening and other operations on the input digitized data to generate intermediate processing results. Storage module: using digital storage media, it is used to store the item processing record data information table RF, the structured data record table and the intermediate data in the operation process, and supports digital reading, writing and indexing of data. Output module: converts the generated processing information into a standardized digital format (such as JSON, XML), and outputs it to other digital systems or terminals through a data interface.

[0054] Furthermore, S200 specifically includes: S210, obtain the processing weight threshold set NS(i) corresponding to the item sub-cluster corresponding to the i-th row of data in the data record table = {N1(i), N2(i), ..., N j (i), ..., N m (i)}, where N j (i) is the maximum processing weight threshold of the item subcluster corresponding to the i-th row of data under the j-th processing method. The value of j ranges from 1 to m, m is the number of processing methods for the item subcluster corresponding to the i-th row of data, and the value of i ranges from 1 to HM, HM is the number of row data in the data record table.

[0055] S220, traverse NS(i), for the traversed N j (i), if N j (i) ≥ N i , N j (i) The unique identifier of the corresponding processing method is added to the processing method record set; N i It is the processing weight in the i-th row of data in the data record table. The initial value of the processing method record set is empty.

[0056] S230, if the number of unique identifiers Q in the processing method record set is 1, the unique identifier in the processing method record set is used as the target identifier corresponding to the i-th row of data, and the processing information of the items to be processed corresponding to the i-th row of data is generated as: single batch processing, the processing method is the processing method corresponding to the target identifier; if Q>1, obtain the fusion value corresponding to each unique identifier in the processing method record, and use the unique identifier corresponding to the minimum fusion value as the target identifier corresponding to the i-th row of data, and generate the processing information of the items to be processed corresponding to the i-th row of data as: single batch processing, the processing method is the processing method corresponding to the target identifier; if Q=0, execute S240.

[0057] S240, obtain the processing batch C(i, j)=roundup(N i / N j (i)), roundup() means rounding up; where the processing weight corresponding to each processing batch in the first C(i, j)-1 processing batches is N j (i), the processing weight corresponding to the C(i, j)th processing batch is N i -(C(i, j)-1)×N j (i).

[0058] S250, obtaining the total energy consumption ET of the item to be processed corresponding to the i-th row of data under the j-th processing method ij .

[0059] ET ij =∑ C（i，j） v=1 E ij (v) + (C (i, j) - 1) × △E0, where E ij (v) is the energy consumption parameter value corresponding to the vth batch, which is determined based on the energy consumption-weight correlation expression corresponding to the jth processing method. Specifically, the processing weight corresponding to the vth batch is substituted into the corresponding energy consumption-weight correlation expression to obtain E ij (v), where v ranges from 1 to C(i, j). ΔE0 is the startup energy consumption between batches, which can be determined based on factors such as the type and power of the processing equipment. For industrial laundry equipment, the instantaneous power consumption during startup can reach 45 kW. With a one-minute startup time, the startup energy consumption is approximately 0.75 kWh. With the use of an inverter, the startup power can be reduced to 30 kW, and the energy consumption per one-minute startup time is approximately 0.5 kWh.

[0060] S260, get min (ET i1 , E.T. i2 ,……,ET ij ,……,ETim ) is used as the target processing method for the i-th row of data, and the processing information for the items to be processed corresponding to the i-th row of data is generated as follows: batch processing, the processing weight of the v-th batch under the j-th processing method is N ij (v), the processing method is the target processing method.

[0061] In the embodiment of the present invention, if min(ET i1 ,ET i2 ,……,ET ij ,……,ET im ) There are multiple corresponding processing methods, and the processing method with the smallest batch size can be selected as the target processing method.

[0062] Through S210 to S260 , the processing information of the items to be processed corresponding to all rows of data in the data record table can be obtained.

[0063] Furthermore, in another embodiment of the present invention, the following steps are also included: S300, set counter h=1.

[0064] S310, if h≤HM, execute S320, otherwise, execute S380.

[0065] S320 , for the hth processing information in the processing information set corresponding to the data record table, if the processing weight of the last batch in the processing information is less than the corresponding maximum processing weight threshold, execute S330 .

[0066] S330, if the intermediate processing information set includes the hth processing information, delete the hth processing information from the intermediate processing information set, and obtain the processing information with the same processing method as the hth processing information, similar visual perception attributes and item type as the processing information corresponding to the hth processing information from the intermediate processing information set with the hth processing information deleted as the candidate information of the hth processing information; if the intermediate processing information set does not include the hth processing information, obtain the processing information with the same processing method as the hth processing information, similar visual perception attributes and item type as the processing information corresponding to the hth processing information, and the processing weight of the last batch is less than the corresponding maximum processing weight threshold from the intermediate processing information set as the candidate information of the hth processing information, and the initial value of the intermediate processing information set is the processing information set corresponding to the data record table.

[0067] In an embodiment of the present invention, similar visual perception attributes refer to that the differences in visual features of the objects are within a preset threshold range, which may specifically include: color: belonging to the same color system (such as light red and pink, dark blue and navy blue), or a color difference value (such as CIEDE2000 color difference value) ≤ 5; texture: the similarity of texture density, direction, and roughness is ≥ 80% (using an image texture feature extraction algorithm, such as a gray level co-occurrence matrix, to calculate the cosine similarity of the feature vector); morphology: the matching degree of geometric shapes is ≥ 70% (such as the overlap of shape contours calculated by an edge detection algorithm between a circle and an ellipse, or a rectangle and a square).

[0068] Similar item types refer to items whose materials, uses, and physical properties belong to adjacent subcategories in the same category system. For example, at the material level, cotton and linen (cotton and flax) and chemical fibers (polyester and nylon) are considered similar; at the use level, kitchen tableware (bowls and plates) and clothing accessories (scarves and shawls) are considered similar; at the physical property level, the difference in parameters such as hardness and temperature resistance is ≤20% (based on parameter grading in industry standards).

[0069] The above determination can be achieved through automatic matching through a preset feature comparison database. When the similarity meets the above threshold, it is determined to be "similar".

[0070] S340, obtaining the sum NT of the processing weight of the h-th processing information and the processing weight of the x-th candidate information of the h-th processing information. hx If the common processing method corresponding to the hth processing information and the xth candidate information has a maximum processing weight threshold greater than or equal to NT hx The processing method is used as the target candidate processing method corresponding to the h-th processing information and the x-th candidate information, and S360 is executed; otherwise, S350 is executed.

[0071] In the embodiment of the present invention, if the common processing method corresponding to the hth processing information and the xth candidate information has a maximum processing weight threshold greater than or equal to NT hx The number of processing methods is multiple, and the processing method with the smallest maximum processing weight threshold is selected as the target candidate processing method corresponding to the h-th processing information and the x-th candidate information.

[0072] S350, set x=x+1, if x≤g(h), execute S340, otherwise execute S370; the initial value of x is 1, g(h) is the number of candidate information for the h-th processing information.

[0073] S360, obtaining the sum SM1 of the fusion value corresponding to the hth processing information and the fusion value corresponding to the xth processing information hx, and the energy consumption weight correlation expression and cleanliness weight correlation expression corresponding to the target candidate processing method corresponding to the hth processing information and the xth candidate information, obtain NT hx Corresponding fusion value SM2 hx , if SM2 hx <SM1 hx , SM2 hx The candidate fusion value is added to the candidate fusion value record set and S350 is executed; otherwise, S350 is executed; the initial value of the candidate fusion value added to the candidate fusion value record set is empty.

[0074] At step S370, the candidate information corresponding to the minimum candidate fusion value in the candidate fusion value record set is used as the target candidate information for the h-th processing information, and merged processing information corresponding to the h-th processing information is generated. The h-th processing information and the corresponding target candidate information are deleted from the intermediate processing information set, h is set to h+1, and step S310 is executed. The merged processing information corresponding to the h-th processing information includes merging the unprocessed items in the last batch of the h-th processing information and the unprocessed items in the last batch of the target candidate information, and processing them using the processing method corresponding to the minimum candidate fusion value. S380 , updating the processing information obtained in S200 based on all the combined processing information, and obtaining updated processing information as final processing information of the data recording table.

[0075] In the embodiment of the present invention, S380 may specifically include: S3801, identifying the original information to be replaced, including locating the original record corresponding to the hth processing information and its target candidate information (including processing batches, weight of each batch, processing method, etc.) from the processing information generated in S200.

[0076] S3802, replace the merged batch information, including deleting the record of "the last batch" in the h-th processing information and the record of "the last batch" in the target candidate information; and adding a new merged batch record: the processing weight is NThx (the sum of the weights of the last batches of the two), the processing method is "the processing method corresponding to the minimum candidate fusion value", and the fusion value is updated to SM2hx.

[0077] S3803: retain the information of the non-merged batches, including the hth processing information and the other batches except the last batch in the target candidate information, and keep their processing weight, processing method and other information unchanged.

[0078] S3804, update the batch number and statistical information, including renumbering all merged batches in sequence (for example, after the original batches 1, 2, and 3 are merged with the candidate batches 1 and 2, the new batch numbers are 1, 2, 3, and 4); and update the total number of processed batches (total batches = original total batches - 2 + 1 = original total batches - 1), and the total fusion value (total fusion value = original total fusion value - SM1hx + SM2hx).

[0079] S3804 generates a final processing information table, integrating the updated batch information, processing method, total energy consumption, total cleanliness, and other parameters into a structured table. This table replaces the original processing information generated in S200 and serves as the final output. For example, the original processing information contains three batches (weights 5kg, 5kg, and 3kg), and the target candidate information contains two batches (weights 5kg and 2kg). After merging, the last two batches (3kg and 2kg) are deleted, and a new merged batch of 5kg (3kg + 2kg) is added. The final batches are 5kg, 5kg, and 5kg, and the processing method is updated to the optimal method corresponding to the merge.

[0080] Through the above steps, it is ensured that the updated processing information not only reflects the results of the merge optimization, but also completely retains the valid data in the original processing logic.

[0081] In the embodiment of the present invention, the following technical effects can be achieved through S300: Improve processing efficiency and reduce energy waste: The aforementioned information is generated only for processing information of a single item subcluster. It is possible that the last batch of processing information in multiple processing data sets may have a smaller weight, leading to frequent equipment startups or low resource utilization. By merging the last batch of items to be processed, the S300 can reduce processing batches and reduce startup energy consumption between batches (e.g., ΔE0). Furthermore, by comparing fusion values ​​and selecting the optimal merging method, the S300 can further reduce overall energy consumption and improve energy efficiency.

[0082] Optimize the flexibility and cost-effectiveness of processing solutions: By identifying candidate information for merging, processing batches can be dynamically adjusted without compromising processing effectiveness, avoiding the inefficient "small batch, multiple batch" processing model. Especially when multiple processing information shares similar attributes, merging processing can reduce equipment idle time, increase processing throughput per unit time, and enhance the cost-effectiveness of the solution.

[0083] Enhanced rationality and accuracy of processed information: During the merging process, fusion values ​​(SM1hx and SM2hx) are compared to ensure that the combined treatment method achieves the best overall performance in terms of energy consumption and cleanliness, rather than blindly merging. Furthermore, clear criteria for selecting candidate information (such as identical treatment methods and similar properties) are clearly defined to ensure the feasibility of merging and ensure that the final processed information better meets actual needs.

[0084] The energy allocation method for item handling provided by this embodiment transforms the energy allocation problem into a data-driven optimization decision by constructing a mapping system between item characteristics, processing methods, and energy consumption models. This method achieves a leap from extensive empirical control to precise algorithmic control. Its core innovation lies in the deep coupling of visual perception technology with energy consumption models, providing a systematic energy-saving solution for high-energy item handling systems (such as laundry and food processing).

[0085] Based on the same inventive concept, an embodiment of the present invention further provides an energy distribution system for article processing, the system comprising: Storage module, the storage module stores the article processing record data information table RF; RF={RF1, RF2, ..., RF u ,……,RF q}, RF u is the record information of the u-th item cluster, where u ranges from 1 to q, and q is the number of item clusters. Item clusters are divided based on the visual perception attributes of items; RF u ={RF u1 ,……,RF uv ,……,RF uf（u）}, RF uv For RF u The record information of the vth item sub-cluster in , where v ranges from 1 to f(u), and f(u) is RF u The number of item subclusters in RF uv ={U uv , RF uv 1 ,……,RF uv r ,……,RF uzv z（u，v）}, U uv is the unique identifier of the vth item sub-cluster, RF uv r is the record information of the v-th item sub-cluster under the r-th processing method, where r ranges from 1 to z(u, v), and z(u, v) is the number of processing methods for the v-th item sub-cluster; RF uv r =(UT uv r , WT uv r , YE uv r , YC uv r ), U uv r is the unique identifier of the rth processing method, WT uv rYE is the maximum processing weight threshold of a single batch of the vth item sub-cluster under the rth processing method, uv r is the energy consumption-weight correlation expression of the vth item sub-cluster under the rth treatment mode, YC uv r is the cleanliness weight association expression of the vth item subcluster under the rth treatment method.

[0086] The data acquisition module is used to obtain the digitized visual feature data and the numerical processing weight data of the current items to be processed through the data acquisition interface to generate a structured data record table for storage in the storage module; wherein each row of data in the data record table corresponds to an item subcluster, and each row of data includes: a unique identifier of the corresponding item subcluster and the processing weight of the items to be processed in the corresponding item subcluster.

[0087] The processing information generation module is used to call the item processing record data information table and the data record table in the storage module, generate processing information for the current item to be processed, and output the processing information, wherein the processing information includes processing mode and processing batch information.

[0088] The system can be used to perform Figure 1 Therefore, for the functions that can be realized by each functional module of the system, please refer to Figure 1 The description of the illustrated embodiment is omitted for brevity.

[0089] An embodiment of the present invention also provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are configured to execute the method described in the embodiment of the present invention.

[0090] An embodiment of the present invention further provides a computer-readable storage medium storing computer-executable instructions, wherein the computer instructions are used to execute the method described in the embodiment of the present invention.

[0091] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in the present invention can be achieved. This is not limited herein.

[0092] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for distributing energy for article processing, characterized in that: The method is implemented based on a data processing system, which includes a storage module. The storage module stores an article processing record data information table RF, wherein RF={RF1, RF2, ..., RF u ,……,RF q }, RF u is the record information of the u-th item cluster, where u ranges from 1 to q, and q is the number of item clusters. Item clusters are divided based on the visual perception attributes of items; RF u ={RF u1 ,……,RF uv ,……,RF uf（u） }, RF uv For RF u The record information of the vth item sub-cluster in , where v ranges from 1 to f(u), and f(u) is RF u The number of item subclusters in RF uv ={U uv , RF uv 1 ,……,RF uv r ,……,RF uzv z（u，v） }, U uv is the unique identifier of the vth item sub-cluster, RF uv r is the record information of the v-th item sub-cluster under the r-th processing method, where r ranges from 1 to z(u, v), and z(u, v) is the number of processing methods for the v-th item sub-cluster; RF uv r =(UT uv r , WT uv r , YE uv r , YC uv r ), U uv r is the unique identifier of the rth processing method, WT uv r YE is the maximum processing weight threshold of a single batch of the vth item sub-cluster under the rth processing method, uv r is the energy consumption-weight correlation expression of the vth item sub-cluster under the rth treatment mode, YC uv r is the cleanliness weight association expression of the vth item subcluster under the rth processing mode; the method comprises the following steps: S100: The data processing system obtains, through a data acquisition interface, digitized visual feature data and numerical processed weight data of the items to be processed, to generate a structured data record table, which is stored in a storage module. Each row of data in the data record table corresponds to an item subcluster, and each row of data includes: a unique identifier of the corresponding item subcluster and the processed weight of the items to be processed in the corresponding item subcluster. S200, the data processing system calls the item processing record data information table and the data record table in the storage module, generates processing information for the current item to be processed, and outputs the processing information, wherein the processing information includes processing mode and processing batch information.

2. The method according to claim 1, characterized in that WT uv r Obtain it in the following ways: S1, obtain the historical data set of the vth item subcluster under the rth processing method, each historical data in the historical data set includes the processing weight, energy consumption parameter value and cleanliness parameter value; S2, deleting the historical data whose cleanliness parameter value is less than the set cleanliness parameter value threshold in the historical data set, and obtaining the deleted historical data set as the target data set; S3, for each data in the target data set, obtain the fusion value corresponding to the data, and the fusion value satisfies the following conditions: SM = We × E + Wc × (Cmax - C), SM is the fusion value corresponding to the data, E is the energy consumption parameter value corresponding to the data, C is the cleanliness parameter value corresponding to the data, We is the energy consumption weight, Wc is the cleanliness weight, We + Wc = 1, and Cmax is the maximum cleanliness parameter value; S4, using the treatment weight in the target data set as the independent variable and the fusion value in the target data set as the dependent variable to obtain the data fitting curve; S5, obtaining the treatment weight corresponding to the slope mutation point of the data fitting curve as WT uv r .

3. The method according to claim 2, characterized in that The energy consumption parameter value E corresponding to the dth historical data d The following conditions are met: d =(Ea d -Eb d ) / Eb d ×100, Ea d is the actual energy consumption value corresponding to the processing weight in the dth historical data, Eb d is the benchmark energy consumption value corresponding to the processing weight in the d-th historical data, d ranges from 1 to Q, and Q is the number of historical data.

4. The method according to claim 3, characterized in that S200 specifically includes: S210, obtain the processing weight threshold set NS(i) corresponding to the item sub-cluster corresponding to the i-th row of data in the data record table = {N1(i), N2(i), ..., N j (i), ..., N m (i)}, where N j (i) is the maximum processing weight threshold of the item subcluster corresponding to the i-th row of data under the j-th processing method. The value of j ranges from 1 to m, m is the number of processing methods of the item subcluster corresponding to the i-th row of data, and the value of i ranges from 1 to HM, HM is the number of row data in the data record table; S220, traverse NS(i), for the traversed N j (i), if N j (i) ≥ N i , N j (i) The unique identifier of the corresponding processing method is added to the processing method record set; N i is the treatment weight in the i-th row of data in the data record table, and the initial value of the treatment method record set is empty; At step S230, if the number of unique identifiers Q in the processing method record set is 1, the unique identifier in the processing method record set is used as the target identifier corresponding to the i-th row of data, and the processing information for the item to be processed corresponding to the i-th row of data is generated as follows: single batch processing, the processing method is the processing method corresponding to the target identifier; if Q is greater than 1, the fusion value corresponding to each unique identifier in the processing method record is obtained, the unique identifier corresponding to the minimum fusion value is used as the target identifier corresponding to the i-th row of data, and the processing information for the item to be processed corresponding to the i-th row of data is generated as follows: single batch processing, the processing method is the processing method corresponding to the target identifier; if Q is 0, proceed to step S240; S240, obtain the processing batch C(i, j)=roundup(N i / N j (i)), roundup() means rounding up; where the processing weight corresponding to each processing batch in the first C(i, j)-1 processing batches is N j (i), the processing weight corresponding to the C(i, j)th processing batch is N i -(C(i, j)-1)×N j (i) S250, obtaining the total energy consumption ET of the item to be processed corresponding to the i-th row of data under the j-th processing method ij ; ET ij =∑ C（i，j） v=1 E ij (v) + (C (i, j) - 1) × △E0, where E ij (v) is the energy consumption parameter value corresponding to the vth batch, which is determined based on the energy consumption weight correlation expression corresponding to the jth treatment method. The value of v ranges from 1 to C(i, j). △E0 is the startup energy consumption between batches. S260, get min (ET i1 ,ET i2 ,……,ET ij ,……,ET im ) is used as the target processing method for the i-th row of data, and the processing information for the items to be processed corresponding to the i-th row of data is generated as follows: batch processing, the processing weight of the v-th batch under the j-th processing method is N ij (v), the processing method is the target processing method.

5. The method according to claim 4, characterized in that The following steps are also included: S300, set counter h=1; S310, if h≤HM, execute S320, otherwise, execute S380; S320, for the hth processing information in the processing information set corresponding to the data record table, if the processing weight of the last batch in the processing information is less than the corresponding maximum processing weight threshold, execute S330; S330: If the intermediate processing information set includes the hth processing information, delete the hth processing information from the intermediate processing information set, and obtain, from the intermediate processing information set from which the hth processing information was deleted, processing information with the same processing method, similar visual perception attributes, and item type as the corresponding hth processing information as the hth processing information as candidate information for the hth processing information. If the intermediate processing information set does not include the hth processing information, obtain, from the intermediate processing information set, processing information with the same processing method, similar visual perception attributes, and item type as the corresponding hth processing information as the hth processing information, and with the processing weight of the last batch less than the corresponding maximum processing weight threshold as candidate information for the hth processing information. The initial value of the intermediate processing information set is the processing information set corresponding to the data record table. S340, obtaining the sum NT of the processing weight of the h-th processing information and the processing weight of the x-th candidate information of the h-th processing information. hx If the common processing method corresponding to the hth processing information and the xth candidate information has a maximum processing weight threshold greater than or equal to NT hx The processing method is used as the target candidate processing method corresponding to the h-th processing information and the x-th candidate information, and S360 is executed; otherwise, S350 is executed; S350, set x = x + 1, if x ≤ g (h), execute S340, otherwise execute S370; the initial value of x is 1, g (h) is the number of candidate information for the h-th processing information; S360, obtaining the sum SM1 of the fusion value corresponding to the hth processing information and the fusion value corresponding to the xth processing information hx , and the energy consumption weight correlation expression and cleanliness weight correlation expression corresponding to the target candidate processing method corresponding to the hth processing information and the xth candidate information, obtain NT hx Corresponding fusion value SM2 hx , if SM2 hx <SM1 hx , SM2 hx Add the candidate fusion value to the candidate fusion value record set and execute S350; Otherwise, directly execute S350; The initial value of the candidate fusion value added to the candidate fusion value record set is empty; At step S370, the candidate information corresponding to the minimum candidate fusion value in the candidate fusion value record set is used as the target candidate information for the h-th processing information, and merged processing information corresponding to the h-th processing information is generated. The h-th processing information and the corresponding target candidate information are deleted from the intermediate processing information set, h=h+1 is set, and step S310 is executed. The merged processing information corresponding to the h-th processing information includes merging the unprocessed items in the last batch of the h-th processing information and the unprocessed items in the last batch of the target candidate information, and processing them using the processing method corresponding to the minimum candidate fusion value. S380 , updating the processing information obtained in S200 based on all the combined processing information, and obtaining updated processing information as final processing information of the data recording table.

6. The method according to claim 1, characterized in that The visual perception attribute is color, and the item type includes size, material, and dirtiness parameter value.

7. An electronic device, characterized in that: including processor and memory; The processor is configured to execute the steps of the method according to any one of claims 1 to 6 by calling the program or instructions stored in the memory.

8. An energy distribution system for article processing, characterized in that: The system comprises: Storage module, the storage module stores the article processing record data information table RF; RF={RF1, RF2, ..., RF u ,……,RF q }, RF u is the record information of the u-th item cluster, where u ranges from 1 to q, and q is the number of item clusters. Item clusters are divided based on the visual perception attributes of items; RF u ={RF u1 ,……,RF uv ,……,RF uf（u） }, RF uv For RF u The record information of the vth item sub-cluster in , where v ranges from 1 to f(u), and f(u) is RF u The number of item subclusters in RF uv ={U uv , RF uv 1 ,……,RF uv r ,……,RF uzv z（u，v） }, U uv is the unique identifier of the vth item sub-cluster, RF uv r is the record information of the v-th item sub-cluster under the r-th processing method, where r ranges from 1 to z(u, v), and z(u, v) is the number of processing methods for the v-th item sub-cluster; RF uv r =(UT uv r , WT uv r , YE uv r , YC uv r ), U uv r is the unique identifier of the rth processing method, WT uv r YE is the maximum processing weight threshold of a single batch of the vth item sub-cluster under the rth processing method, uv r is the energy consumption-weight correlation expression of the vth item sub-cluster under the rth treatment mode, YC uv r is the cleanliness weight association expression of the vth item subcluster under the rth treatment method; A data acquisition module, configured to acquire, through a data acquisition interface, digitized visual feature data and numerical processed weight data of the items to be processed, to generate a structured data record table for storage in the storage module; wherein each row of data in the data record table corresponds to an item subcluster, and each row of data includes: a unique identifier of the corresponding item subcluster and the processed weight of the items to be processed in the corresponding item subcluster; The processing information generation module is used to call the item processing record data information table and the data record table in the storage module, generate processing information for the current item to be processed, and output the processing information, wherein the processing information includes processing mode and processing batch information.

Images