A scenario-based adaptive sorting decision method, system and sorting equipment
By constructing a sorting scenario matrix and adaptive decision-making method, combining cargo characteristics and sorting device parameters, the problems of low efficiency and poor adaptability of automated sorting equipment in the prior art are solved, and efficient and accurate sorting operations are achieved.
Patent Information
- Application Number
- CN202411205221.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-08-30
AI Technical Summary
The existing automated sorting equipment design is designed to meet specific needs, and it is difficult to achieve highly intelligent decision-making, and it is difficult to quickly adapt to changes in sorting demands in different scenarios, resulting in inefficiency and high error rates.
By building a sorting scenario matrix, combining cargo specifications, surface friction characteristics and vulnerable parameters, we quickly generate the optimal sorting strategy, select an appropriate balance wheel or cross-belt sorting device for sorting, and realize adaptive decision-making.
It improves sorting efficiency, reduces error rate, can quickly adapt to changes in sorting requirements in different scenarios, and improves the accuracy and adaptability of sorting.
Smart Images

Figure CN118698892B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and specifically to a scenario-based adaptive sorting decision method, system and sorting equipment. Background Art
[0002] In the logistics, manufacturing, retail and other industries, sorting operations are a key link, and their efficiency and quality directly affect the performance of the entire supply chain. With the rapid development of technologies such as the Internet of Things, big data, and artificial intelligence, scenario-based adaptive sorting decision methods, systems, and sorting equipment have become a hot topic of research. Traditional sorting methods mainly rely on manual or simple automated equipment, and have problems such as low sorting efficiency, high error rate, and high labor costs. At the same time, due to differences in sorting requirements in different scenarios, such as the type, size, weight, and shape of the sorted items, traditional sorting methods often find it difficult to meet these diverse needs.
[0003] In summary, the existing technology relies on simple automation equipment, and most of the existing automation equipment is designed for specific sorting needs, resulting in the inability to achieve highly intelligent sorting decisions and difficulty in quickly adapting to changes in sorting needs in different scenarios. Summary of the invention
[0004] Based on this, it is necessary to provide a scenario-based adaptive sorting decision method, system and sorting equipment to address the above technical problems. By constructing a sorting scenario matrix, the optimal sorting strategy can be quickly generated to improve the sorting efficiency and achieve the technical effect of quickly adapting to changes in sorting requirements in different scenarios.
[0005] In a first aspect, a scenario-based adaptive sorting decision method is provided, the method comprising: acquiring a batch of goods to be sorted, collecting sorting categories for the batch of goods, and obtaining the number of sorting categories; acquiring images of multiple sorted goods in the batch of goods, and performing goods specification recognition based on the multiple goods images to obtain multiple goods specification parameters; performing goods surface friction feature recognition based on the multiple goods images to obtain multiple surface friction parameters; acquiring goods categories of the multiple sorted goods, performing vulnerability analysis based on the multiple goods categories to obtain multiple vulnerability parameters; and determining the number of sorted goods according to the multiple surface friction parameters, the multiple goods specification parameters, and the multiple According to the pre-built pendulum wheel sorting parameters and cross-belt sorting parameters of the pendulum wheel sorting device and the cross-belt sorting device, a sorting scenario matrix is constructed, and a sorting adaptability analysis is performed to obtain a pendulum wheel adaptability matrix and a cross-belt adaptability matrix; according to the number of sorting categories, a first sorting decision is made to obtain a first decision probability distribution, and according to the pendulum wheel adaptability matrix and the cross-belt adaptability matrix, a second sorting decision is made to obtain a second decision probability distribution, and a sorting decision probability distribution is calculated in combination with the first decision probability distribution, and a pendulum wheel sorting device or a cross-belt sorting device is selected according to the sorting decision probability distribution to sort the batch of goods.
[0006] In a second aspect, a scenario-based adaptive sorting decision system is provided, the system comprising: a sorting category acquisition module, the sorting category acquisition module is used to acquire a batch of goods to be sorted, perform sorting category acquisition on the batch of goods, and obtain the number of sorting categories; a goods specification parameter acquisition module, the goods specification parameter acquisition module is used to acquire images of a plurality of sorted goods in the batch of goods, perform goods specification recognition based on the plurality of goods images, and obtain a plurality of goods specification parameters; a surface friction parameter acquisition module, the surface friction parameter acquisition module is used to perform goods surface friction feature recognition based on the plurality of goods images, and obtain a plurality of surface friction parameters; a goods category acquisition module, the goods category acquisition module is used to acquire the goods categories of the plurality of sorted goods, perform vulnerability analysis based on the plurality of goods categories, and obtain a plurality of vulnerability parameters; An adaptability analysis module, the sorting adaptability analysis module is used to construct a sorting scenario matrix according to the multiple surface friction parameters, multiple cargo specification parameters and multiple damage parameters, and perform sorting adaptability analysis according to the pre-constructed balance wheel sorting parameters and cross belt sorting parameters of the balance wheel sorting device and the cross belt sorting device to obtain the balance wheel adaptability matrix and the cross belt adaptability matrix; a sorting decision probability calculation module, the sorting decision probability calculation module is used to make a first sorting decision according to the number of sorting categories to obtain a first decision probability distribution, make a second sorting decision according to the balance wheel adaptability matrix and the cross belt adaptability matrix to obtain a second decision probability distribution, calculate and obtain the sorting decision probability distribution in combination with the first decision probability distribution, and select the balance wheel sorting device or the cross belt sorting device according to the sorting decision probability distribution to sort the batch of goods.
[0007] According to a third aspect, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps according to the first aspect when executing the computer program.
[0008] The above-mentioned scenario-based adaptive sorting decision method, system and sorting equipment solve the technical problems that the existing technology relies on simple automated equipment, and most of the existing automated equipment is designed for specific sorting needs, resulting in the inability to achieve highly intelligent sorting decisions and difficulty in quickly adapting to changes in sorting needs in different scenarios. By constructing a sorting scenario matrix, the optimal sorting strategy is quickly generated, the sorting efficiency is improved, and the technical effect of quickly adapting to changes in sorting needs in different scenarios is achieved.
[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] Figure 1 A schematic diagram of a flow chart of a scenario-based adaptive sorting decision method in an embodiment;
[0011] Figure 2 A schematic diagram of a flow chart of obtaining a vulnerable function of a scenario-based adaptive sorting decision method in one embodiment;
[0012] Figure 3 The structure block diagram of a scenario-based adaptive sorting decision system in one embodiment.
[0013] Explanation of the reference numerals: sorting category acquisition module 11, cargo specification parameter acquisition module 12, surface friction parameter acquisition module 13, cargo category acquisition module 14, sorting adaptability analysis module 15, sorting decision probability calculation module 16. DETAILED DESCRIPTION
[0014] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0015] like Figure 1 As shown, the present application provides a scenario-based adaptive sorting decision method, the method comprising:
[0016] Acquire a batch of goods to be sorted, collect sorting categories for the batch of goods, and obtain the number of sorting categories.
[0017] Scenario-based sorting is a strategy that combines services, products, content or experiences with specific environments or situations to create a more personalized, relevant and engaging user experience. In this application, the scenario is the decision to use a pendulum wheel sorting device or a cross-belt sorting device for a batch of goods. The pendulum wheel sorting device requires stronger surface friction of the goods, and the cross-belt sorting device moves the goods quickly during sorting, and the probability of fragile goods being damaged is high. Both need to consider the specifications of the goods. The adaptive sorting decision method is an intelligent sorting strategy that automatically adjusts and optimizes the sorting strategy based on real-time logistics scenarios, cargo characteristics, sorting targets and other factors. The core lies in its "adaptive" feature, that is, the ability to flexibly adjust the sorting strategy according to changes in the environment and conditions to achieve more efficient, accurate and flexible sorting operations. The present application provides a scenario-based adaptive sorting decision method, which collects sorting categories of batches of goods to be sorted, obtains the number of sorting categories, and adaptively adjusts the sorting strategy according to the real-time logistics scenario and goods characteristics. Therefore, the sorting operation can be guided more accurately, and unnecessary movement and waiting time can be reduced, thereby significantly improving the sorting efficiency. The scenario-based adaptive sorting decision method can select the optimal sorting path and strategy based on real-time data and algorithm analysis, thereby reducing the sorting error rate and improving the sorting accuracy.
[0018] Obtain batches of goods to be sorted from sources such as warehouses or production lines. To be sorted refers to items or data that have not yet started or are waiting for some operation or processing. In this application, it refers to goods waiting to be sorted. In the field of logistics or production, batch goods usually combine multiple similar items or products together and process or transport them as a unit. In this application, it refers to a group of goods waiting to be sorted. Sorting category collection includes identifying and classifying goods to determine which category they should be sorted into, which refers to identifying and classifying goods waiting to be sorted to determine their destination or other related attributes. The number of sorting categories refers to the number or quantity statistics of different categories after classification, and in this method, it refers to the number of each category after all goods are classified. For example, the goods are classified according to the destination, and it is assumed that there are five different destination categories. Through the above method, the sorting category collection of the batch goods to be sorted and the number of sorting categories are obtained can improve the sorting efficiency and provide a basis for the follow-up.
[0019] According to the sorting identifiers of the multiple sorted goods in the batch of goods, multiple sorting directions of the multiple sorted goods are extracted; the multiple sorting directions are classified to obtain the number of sorting categories.
[0020] According to each sorted cargo in the batch of cargo, it is necessary to identify the sorting mark thereon, including a barcode, an RFID tag, a text description or any method that can identify the attributes of the cargo. The images of multiple sorted cargo in the batch of cargo are collected from the sorting mark, and the cargo specifications are identified according to the multiple cargo images to obtain multiple cargo specification parameters; and the information of the sorting direction is extracted. The sorting direction usually refers to the destination where the cargo needs to arrive at or the distribution path that needs to be followed. This can be a specific geographical location, such as a city, a warehouse address, etc., or it can be a certain logical classification, such as a product type, a customer type, etc. The results of the classification and calculation are verified to ensure the accuracy and completeness of the data. Then, the results are stored in a database or an appropriate storage system for subsequent analysis and query. Through this process, the sorting requirements of the cargo in each batch can be more accurately grasped, so that logistics planning and operation can be carried out more efficiently. At the same time, this also provides important data support for subsequent data analysis, decision making, etc.
[0021] The method comprises collecting images of a plurality of sorted goods in the batch of goods to obtain a plurality of goods images; obtaining a set of sample goods images, and marking the goods specification and size parameters in the sample goods images to obtain a set of sample goods specification parameters; constructing a goods specification identifier using the set of sample goods images and the set of sample goods specification parameters; and performing goods specification identification on the plurality of goods images based on the goods specification identifier to obtain the plurality of goods specification parameters.
[0022] Use image acquisition equipment, such as cameras, scanners, etc., to acquire images of multiple sorted goods in the batch of goods. The image acquisition equipment can capture visual information such as the appearance, shape and size of the goods, generate corresponding goods images, that is, obtain multiple goods images; prepare a set of sample goods images with known specifications and sizes. This set of images should cover goods of different specifications, sizes and shapes so that the goods specification identifier can learn enough features. For each image in the sample goods image set, it is necessary to mark the goods specification and size parameters therein, including the length, width, height and other size information of the goods, as well as other specification parameters that may affect the sorting, such as weight, volume, etc. The marking process can be completed by manual marking or using a special marking tool. Using machine learning or deep learning technology, a goods specification identifier is trained using a sample goods image set and a corresponding sample goods specification parameter set. The goods specification identifier will learn how to extract features from the goods image and predict the specification and size parameters of the goods based on these features. After the goods specification identifier is trained, it can be applied to the actual sorting scenario. The previously collected multiple cargo images are input into the identifier, which will automatically identify the specifications of the cargo in each image and output the corresponding specification size parameters. Through the processing of the cargo specification identifier, the specification size parameters of multiple cargoes are obtained. The specification size parameters can be used in subsequent cargo sorting, route planning, warehouse management and other links to improve the efficiency and accuracy of the entire logistics system.
[0023] Cargo surface friction features are identified based on the multiple cargo images to obtain multiple surface friction parameters.
[0024] According to the multiple cargo images, preprocessing is performed to eliminate noise in the image, improve contrast and possibly perform image enhancement. The above steps are crucial for subsequent feature recognition because they can improve the clarity of key information in the image. The surface friction characteristics of the cargo are usually closely related to its surface texture. Therefore, it is necessary to extract the surface texture information of the cargo from the image. This can be achieved by using various image filters and texture analysis techniques, such as Gabor filters or Fourier analysis. The extracted texture features may be very large and contain a lot of redundant and irrelevant information. Therefore, feature selection and dimensionality reduction are required to select the most important features and reduce the amount of calculation, which is achieved by using machine learning techniques such as principal component analysis, linear discriminant analysis or random forests. The above techniques can evaluate the importance of each feature and select the most representative feature subset. Therefore, feature selection and dimensionality reduction are required to select the most important features and reduce the amount of calculation. Based on the extracted surface texture features and known surface friction data, if any, a surface friction model can be established, the purpose of which is to predict the surface friction parameters of the cargo based on its surface texture features. The preprocessed cargo image is input into the established surface friction model, and the model will automatically identify the surface friction parameters of the cargo in the image and output the corresponding friction parameter values, which may include static friction coefficient, dynamic friction coefficient, etc., depending on the application scenario and requirements. Through the above method, the surface friction characteristics of the cargo are identified and the surface friction parameters are obtained based on multiple cargo images, which lays a foundation for improving the accuracy and efficiency of cargo sorting in the future.
[0025] The multiple cargo images are cropped to obtain multiple cargo partial images; a set of sample cargo partial images is collected, and cargo surface friction parameters in the sample cargo partial images are marked to obtain a set of sample surface friction parameters; a surface friction identifier is constructed using the set of sample cargo partial images and the set of sample surface friction parameters; the multiple cargo partial images are identified using the surface friction identifier to obtain the multiple surface friction parameters.
[0026] The multiple cargo images collected originally are cropped. The purpose of cropping is to obtain a local image of each cargo, which may contain information related to the surface friction characteristics, such as specific surface texture, material, etc. The cropping process can be based on a fixed area of the cargo, such as a specific part or mark, or automatically determined based on image analysis technology. After cropping, multiple cargo local images are obtained. These images will serve as the basis for subsequent surface friction feature extraction and recognition. In order to train the surface friction identifier, it is necessary to prepare a set of sample cargo local images with known surface friction parameters. This set of images should cover different surface textures, materials and possible friction conditions, and is recorded as a sample cargo local image set. For each image in the sample cargo local image set, it is necessary to identify the cargo surface friction parameters therein, including friction coefficient, friction type, etc. The surface friction parameters of all identified sample cargo local images are integrated into a set, namely, the sample surface friction parameter set, which will be used as the standard for training the surface friction identifier. A surface friction identifier is trained using the sample cargo local image set and the corresponding sample surface friction parameter set using deep learning technology. During the training process, the identifier will learn to extract features related to surface friction from the cargo local images, optimize the model parameters through the back propagation algorithm and the gradient descent method to minimize the error between the predicted value and the true value, and establish a mapping relationship between these features and the surface friction parameters. The trained model is evaluated using the validation set data, and the hyperparameters or structure of the model are adjusted according to the evaluation results to improve the performance of the model, find the optimal hyperparameter combination, and obtain the surface friction identifier. After the surface friction identifier is trained, it is applied to actual local images of goods. The multiple local images of goods that have been cropped before are input into the identifier. The identifier will automatically identify the surface friction parameters of the local goods in each image and output the corresponding friction parameter values.
[0027] The cargo categories of the plurality of sorted cargoes are collected, and vulnerability analysis is performed according to the plurality of cargo categories to obtain a plurality of vulnerability parameters.
[0028] Collect the category information of each sorted cargo from the existing database, logistics system or manual input, classify and organize the collected cargo categories for subsequent analysis, define vulnerability indicators, and determine key indicators for evaluating the vulnerability of cargo, such as material, weight, size, shape, packaging, etc., to quantify the vulnerability of cargo. According to the defined vulnerability indicators, conduct vulnerability analysis on each cargo category, and quantify the analysis results into specific vulnerability parameters, such as vulnerability level, damage probability, etc. The vulnerability parameters will be used for subsequent logistics operations and management decisions. Through the above method, data support on cargo vulnerability is provided to the management to assist in the formulation of logistics strategies, inventory management decisions, etc. By analyzing the vulnerability parameters, the damage risk of cargo can be better predicted and managed.
[0029] like Figure 2 As shown, the cargo categories of the multiple sorted cargoes are collected; a sample cargo category set is obtained, and a vulnerability level classification is performed to obtain a sample vulnerability parameter set; based on a decision tree, a vulnerability classifier is constructed using the sample cargo category set and the sample vulnerability parameter set, and a vulnerability analysis is performed on multiple cargo categories to obtain multiple vulnerability parameters.
[0030] Collect the category information of each sorted cargo from the logistics system, warehouse management system or manual input to ensure the accuracy and completeness of the data, including cargo name, category code, attributes, etc.; select representative sample cargo categories from all collected cargo categories. The samples should be diverse to cover cargo with different degrees of vulnerability as a set of sample cargo categories; classify the vulnerability level of the sample cargo categories according to professional knowledge, historical data or expert evaluation. For example, the vulnerability level can be discrete, such as 1-5, or continuous, such as vulnerability score. Organize the vulnerability level or score of each sample cargo category into a set of sample vulnerability parameters. Select an appropriate decision tree algorithm. The decision tree is a commonly used supervised learning algorithm that constructs a tree structure by recursively dividing the data set into several subsets. Each internal node represents a feature, i.e., an attribute of the goods, each branch represents a possible value of the feature, and each leaf node represents a category, i.e., the degree of vulnerability. The sample vulnerability parameter set refers to assigning a vulnerability parameter to each sample goods category in the sample goods category set, which is usually a numerical value or classification label, used to indicate the degree of vulnerability of the goods category. Before building a decision tree, it is necessary to select features for classification, i.e., important factors that can affect the vulnerability of the goods. For example, for glass products, features such as "size", "packaging method" and "transportation distance" are selected. The decision tree is trained using the sample goods category set and the sample vulnerability parameter set, a root node is selected to represent the entire sample goods category set, and an optimal feature is selected to divide the data set. This optimal feature is usually a feature that can minimize the "impurity" of the data set. Impurity can be measured using indicators such as information gain, gain rate or Gini index. The data set is divided into several subsets according to the different values of the selected features, and a new child node is created for each subset. The above process is repeated recursively for each subset until the stop condition is met, such as the size of the subset is less than a certain threshold, or all subsets belong to the same category. In order to prevent overfitting, the decision tree can be pruned. Pruning can be divided into pre-pruning and post-pruning. Pre-pruning is to stop the growth of the tree in advance during the construction of the decision tree, while post-pruning is to simplify the tree structure by deleting some subtrees or leaf nodes after the tree is built. Obtain the constructed decision tree, that is, the vulnerability classifier, and perform vulnerability analysis on the new cargo category based on the decision tree. The characteristic values of the new cargo are input into the vulnerability classifier, and then the tree structure is traversed from the root node layer by layer until a leaf node is reached. The category represented by the leaf node is the vulnerability parameter of the new cargo. Through the above method, a vulnerability classifier is constructed and multiple vulnerability parameters are obtained, which provides data support for subsequent research.
[0031] A sorting scenario matrix is constructed according to the multiple surface friction parameters, multiple cargo specification parameters and multiple vulnerability parameters, and a sorting adaptability analysis is performed according to the pre-constructed balance wheel sorting parameters and cross belt sorting parameters of the balance wheel sorting device and the cross belt sorting device to obtain a balance wheel adaptability matrix and a cross belt adaptability matrix.
[0032] It is necessary to combine multiple surface friction parameters, multiple cargo specification parameters and multiple fragile parameters into a multi-dimensional sorting scenario matrix. Each row of the sorting scenario matrix can represent a specific sorting scenario, including information such as the friction coefficient between the cargo and the conveyor belt, cargo specification parameters, etc. The pre-built balance wheel sorting device and cross belt sorting device have their own balance wheel sorting parameters and cross belt sorting parameters, including the speed, angle, and stroke of the balance wheel, and the speed, width, and material of the cross belt. These parameters will affect the sorting device's handling capacity and accuracy of the cargo, as well as the degree of damage to the cargo during the sorting process. For each sorting scenario, that is, each row in the sorting scenario matrix, it is necessary to analyze whether the balance wheel sorting device can adapt to this scenario. It is necessary to consider whether the cargo specification parameters, such as size, weight, etc., are within the processing capacity of the balance wheel sorting device, and whether the cargo fragile parameters allow the use of balance wheel sorting, that is, whether the cargo is easily damaged during the sorting process. Based on these analyses, a balance wheel adaptability score is assigned to each sorting scenario to form a balance wheel adaptability matrix. Similarly, for each sorting scenario, to analyze whether the cross-belt sorting device can adapt to this scenario, it is necessary to consider the specifications of the goods, such as whether the size and shape are suitable for transmission on the cross-belt, and whether the fragility parameters of the goods allow the use of cross-belt sorting. Based on these analyses, a cross-belt adaptability score is assigned to each sorting scenario to form a cross-belt adaptability matrix. Through the pendulum adaptability matrix and the cross-belt adaptability matrix, it can be clearly seen which sorting device is more suitable for handling which type of goods. Based on this information, a more reasonable sorting strategy can be formulated to improve sorting efficiency and accuracy while reducing the damage rate of goods.
[0033] Taking surface friction parameters, cargo specification parameters and vulnerability parameters as row attributes and multiple sorted cargoes as column attributes, a sorting scenario matrix is constructed according to the multiple surface friction parameters, multiple cargo specification parameters and multiple vulnerability parameters; a balance wheel sorting adaptability analyzer is constructed, wherein the balance wheel sorting adaptability analyzer includes a balance wheel friction adaptability analysis branch, a balance wheel specification adaptability analysis branch and a balance wheel vulnerability adaptability analysis branch, which respectively take sample balance wheel sorting parameters combined with sample surface friction parameters, sample cargo specification parameters and sample vulnerability parameters as inputs, and take sample balance wheel friction adaptability parameters, sample balance wheel specification adaptability parameters and sample balance wheel vulnerability adaptability parameters as outputs; using the balance wheel sorting adaptability analyzer, the balance wheel sorting parameters are combined with the sorting scenario matrix to perform a balance wheel sorting adaptability analysis to obtain a balance wheel adaptability matrix; constructing a cross-belt sorting adaptability analyzer, the cross-belt sorting parameters are combined with the sorting scenario matrix to perform a cross-belt sorting adaptability analysis to obtain a cross-belt adaptability matrix.
[0034] The row attributes include surface friction parameters, cargo specification parameters and vulnerability parameters. Each row represents a specific sorting scenario and contains the specific values of these parameters. The column attributes refer to multiple sorted cargoes. Each column represents a cargo. The corresponding cargoes in the sorting scenario matrix are the cargoes to which the parameters in each row apply. Constructing the sorting scenario matrix. Constructing the balance wheel sorting adaptability analyzer includes three main analysis branches: the balance wheel friction adaptability analysis branch, which inputs the sample balance wheel sorting parameters and the sample surface friction parameters, and outputs the sample balance wheel friction adaptability parameters, which reflects the adaptability of the balance wheel sorting device under given surface friction conditions. The balance wheel specification adaptability analysis branch, which inputs the sample balance wheel sorting parameters and the sample cargo specification parameters, and outputs the sample balance wheel specification adaptability parameters, which reflects the adaptability of the balance wheel sorting device under given cargo specifications. The balance wheel vulnerability adaptability analysis branch, which inputs the sample balance wheel sorting parameters and the sample vulnerability parameters, and outputs the sample balance wheel vulnerability adaptability parameters, which reflects the adaptability of the balance wheel sorting device when handling vulnerable cargo. Use the constructed balance wheel sorting adaptability analyzer, and combine each row in the sorting scenario matrix, that is, each specific sorting scenario, to perform balance wheel friction adaptability analysis, balance wheel specification adaptability analysis, and balance wheel fragility adaptability analysis. Combine the output results of each analysis branch into a comprehensive balance wheel adaptability score and fill it into the corresponding position of the balance wheel adaptability matrix. Each row of this matrix corresponds to each row of the sorting scenario matrix, and each column corresponds to a specific balance wheel sorting parameter combination. Construct a cross-belt sorting adaptability analyzer. For the cross-belt sorting adaptability analyzer, although there is no subdivided analysis branch like the balance wheel, the basic analysis logic is similar. Input the cross-belt sorting parameters and the sorting scenario matrix, and output the cross-belt adaptability matrix. Each row corresponds to each row of the sorting scenario matrix, and each column corresponds to a specific cross-belt sorting parameter combination. Use the cross-belt sorting adaptability analyzer, and combine each row in the sorting scenario matrix, that is, each specific sorting scenario, to perform cross-belt sorting adaptability analysis, and fill the analysis results into the corresponding position of the cross-belt adaptability matrix. Constructing a sorting scenario matrix as well as a pendulum wheel and cross-belt sorting adaptation analyzer, and combining these analyzers for adaptive analysis, can achieve the technical effect of significantly improving sorting efficiency, reducing cargo damage, and optimizing resource allocation.
[0035] According to the number of sorting categories, a first sorting decision is made to obtain a first decision probability distribution; according to the balance wheel adaptability matrix and the cross belt adaptability matrix, a second sorting decision is made to obtain a second decision probability distribution; in combination with the first decision probability distribution, a sorting decision probability distribution is calculated to obtain a sorting decision probability distribution; and according to the sorting decision probability distribution, a balance wheel sorting device or a cross belt sorting device is selected to sort the batch of goods.
[0036] The first sorting decision is to make a preliminary sorting decision based on the number of sorting categories of the batch of goods, analyze the relationship between the number of sorting categories and the efficiency of the sorting equipment, and obtain a first decision probability distribution, which represents the probability of selecting a balance wheel sorting device or a cross belt sorting device according to the number of sorting categories. The second sorting decision uses the balance wheel adaptability matrix and the cross belt adaptability matrix, combined with the specific parameters of the batch of goods, such as surface friction parameters, cargo specification parameters and fragility parameters, to conduct a more detailed adaptability analysis, considering the adaptability of the balance wheel sorting device and the cross belt sorting device under different parameters, and output a second decision probability distribution, which represents the probability of selecting a balance wheel sorting device or a cross belt sorting device under given parameters. The first decision probability distribution and the second decision probability distribution are integrated. This can be achieved by weighted averaging, Bayesian updating or other suitable probability fusion methods. The result of the integration is a comprehensive sorting decision probability distribution, which comprehensively considers the impact of the number of sorting categories and specific parameters on the selection of sorting equipment. According to the integrated sorting decision probability distribution, a specific sorting device is selected for sorting. If the probability of the balance wheel sorting device is higher than that of the cross belt sorting device, the balance wheel sorting device is selected; otherwise, the cross belt sorting device is selected. For example, a threshold can be set to decide whether to make a deterministic selection, that is, to select the device with the highest probability, or to make a random selection, that is, to make a random selection according to the probability distribution. The above method can improve sorting efficiency and reduce the damage rate of goods by comprehensively considering multiple factors for sorting decisions.
[0037] According to the balance wheel sorting parameters and the cross belt sorting parameters, the number of balance wheel sorting categories and the number of cross belt sorting categories are extracted; the occupancy saturation of the number of sorting categories and the number of balance wheel sorting categories and the number of cross belt sorting categories are calculated to obtain the balance wheel sorting saturation and the cross belt sorting saturation, and the first decision probability distribution is obtained by allocation and calculation; according to the balance wheel adaptability matrix and the cross belt adaptability matrix, the balance wheel adaptability parameters and the cross belt adaptability parameters are calculated; according to the balance wheel adaptability parameters and the cross belt adaptability parameters, the second decision probability distribution is allocated and calculated; according to the first decision probability distribution and the second decision probability distribution, the sorting decision probability distribution is calculated; according to the sorting decision probability distribution, a decision is made to select a balance wheel sorting device or a cross belt sorting device to sort the batch of goods.
[0038] According to the balance wheel sorting parameters and the cross belt sorting parameters, the number of sorting categories that the balance wheel sorting device and the cross belt sorting device can handle are extracted respectively, that is, the number of balance wheel sorting categories and the number of cross belt sorting categories, and the number of sorting categories of the current batch of goods to be sorted is calculated. According to the number of balance wheel sorting categories and the number of cross belt sorting categories, the occupancy saturation of the balance wheel sorting device and the cross belt sorting device is calculated respectively. Balance wheel sorting saturation = current number of balance wheel sorting categories / maximum number of categories of balance wheel sorting device; cross belt sorting saturation = current number of cross belt sorting categories / maximum number of categories of cross belt sorting device; according to the balance wheel sorting saturation and the cross belt sorting saturation, the first decision probability distribution is allocated and calculated. For example, if the saturation of the balance wheel sorting device is low, the balance wheel sorting device is given a higher selection probability. According to the balance wheel adaptability matrix and the cross belt adaptability matrix, the adaptability parameters of the balance wheel sorting device and the cross belt sorting device for the current batch of goods are calculated, and the evaluation is based on factors such as the fragility, specifications, and shape of the goods. According to the balance wheel adaptability parameter and the cross belt adaptability parameter, the second decision probability distribution is allocated and calculated. If a certain sorting device has a higher adaptability to the goods, a higher selection probability is given to the device. Combining the first decision probability distribution and the second decision probability distribution, the final sorting decision probability distribution is calculated by weighting or other methods. According to the sorting decision probability distribution, a balance wheel sorting device or a cross belt sorting device is selected to sort the batch of goods. If the probability of the balance wheel sorting device is higher than that of the cross belt sorting device, the balance wheel sorting device is selected; otherwise, the cross belt sorting device is selected. The selected sorting device is used to sort the batch of goods. Through the above method, multiple factors such as the characteristics of the goods, the capabilities and adaptability of the sorting device are combined, and decisions are made through probability distribution, aiming to achieve more efficient, safer and more economical sorting operations.
[0039] like Figure 3 As shown, the embodiment of the present application includes a scenario-based adaptive sorting decision system, and the system includes:
[0040] A sorting category collection module 11, the sorting category collection module 11 is used to obtain a batch of goods to be sorted, collect the sorting categories of the batch of goods, and obtain the number of sorting categories;
[0041] A goods specification parameter acquisition module 12, the goods specification parameter acquisition module 12 is used to collect images of a plurality of sorted goods in the batch of goods, perform goods specification recognition based on the plurality of goods images, and obtain a plurality of goods specification parameters;
[0042] A surface friction parameter acquisition module 13, wherein the surface friction parameter acquisition module 13 is used to identify the surface friction characteristics of the cargo according to the multiple cargo images and obtain multiple surface friction parameters;
[0043] A cargo category collection module 14, the cargo category collection module 14 is used to collect cargo categories of the plurality of sorted cargoes, perform vulnerability analysis according to the plurality of cargo categories, and obtain a plurality of vulnerability parameters;
[0044] A sorting adaptability analysis module 15, which is used to construct a sorting scenario matrix according to the plurality of surface friction parameters, the plurality of cargo specification parameters and the plurality of vulnerability parameters, and to perform sorting adaptability analysis according to the pre-constructed balance wheel sorting parameters and cross belt sorting parameters of the balance wheel sorting device and the cross belt sorting device, so as to obtain a balance wheel adaptability matrix and a cross belt adaptability matrix;
[0045] A sorting decision probability calculation module 16 is used to make a first sorting decision according to the number of sorting categories to obtain a first decision probability distribution, make a second sorting decision according to the balance wheel adaptability matrix and the cross belt adaptability matrix to obtain a second decision probability distribution, calculate and obtain the sorting decision probability distribution in combination with the first decision probability distribution, and select a balance wheel sorting device or a cross belt sorting device to sort the batch of goods according to the sorting decision probability distribution.
[0046] Furthermore, the embodiment of the present application also includes:
[0047] A sorting direction extraction module, the sorting direction extraction module is used to extract multiple sorting directions of multiple sorted goods according to the sorting identifiers of multiple sorted goods in the batch of goods;
[0048] A classification module is used to classify the multiple sorting directions and obtain the number of sorting categories.
[0049] Furthermore, the embodiment of the present application also includes:
[0050] A cargo image acquisition module, the cargo image acquisition module is used to collect images of a plurality of sorted cargoes in the batch of cargoes to obtain a plurality of cargo images;
[0051] A sample cargo specification parameter set acquisition module, the sample cargo specification parameter set acquisition module is used to acquire a sample cargo image set, and mark cargo specification size parameters in the sample cargo image to acquire a sample cargo specification parameter set;
[0052] A cargo specification identifier construction module, the cargo specification identifier construction module is used to construct a cargo specification identifier using the sample cargo image set and the sample cargo specification parameter set;
[0053] A cargo specification recognition module is used to perform cargo specification recognition on the multiple cargo images based on the cargo specification identifier to obtain the multiple cargo specification parameters.
[0054] Furthermore, the embodiment of the present application also includes:
[0055] A cargo partial image acquisition module, the cargo partial image acquisition module is used to crop the multiple cargo images to obtain multiple cargo partial images;
[0056] The identification module is used to collect a set of local images of sample goods, identify the friction parameters of the goods surface in the local images of the sample goods, and obtain the set of friction parameters of the sample surface;
[0057] A surface friction identifier construction module, wherein the surface friction identifier construction module is used to construct a surface friction identifier using the sample cargo local image set and the sample surface friction parameter set;
[0058] A local image recognition module is used to use the surface friction identifier to recognize the multiple local images of the goods and obtain the multiple surface friction parameters.
[0059] Furthermore, the embodiment of the present application also includes:
[0060] A cargo category collection module, the cargo category collection module is used to collect the cargo categories of the plurality of sorted cargoes;
[0061] A vulnerability level classification module, the vulnerability level classification module is used to obtain a sample cargo category set, perform vulnerability level classification, and obtain a sample vulnerability parameter set;
[0062] The vulnerability analysis module is used to construct a vulnerability classifier based on a decision tree, using the sample cargo category set and the sample vulnerability parameter set, perform vulnerability analysis on multiple cargo categories, and obtain multiple vulnerability parameters.
[0063] Furthermore, the embodiment of the present application also includes:
[0064] A sorting scenario matrix construction module, wherein the sorting scenario matrix construction module is used to construct a sorting scenario matrix according to the surface friction parameters, the cargo specification parameters and the vulnerability parameters, with the surface friction parameters, the cargo specification parameters and the vulnerability parameters as row attributes and the plurality of sorted cargoes as column attributes;
[0065] A balance wheel sorting adaptability analyzer construction module, the balance wheel sorting adaptability analyzer construction module is used to construct a balance wheel sorting adaptability analyzer, wherein the balance wheel sorting adaptability analyzer includes a balance wheel friction adaptability analysis branch, a balance wheel specification adaptability analysis branch and a balance wheel fragility adaptability analysis branch, which are respectively input with a sample balance wheel sorting parameter combined with a sample surface friction parameter, a sample cargo specification parameter and a sample fragility parameter, and are constructed with a sample balance wheel friction adaptability parameter, a sample balance wheel specification adaptability parameter and a sample balance wheel fragility adaptability parameter as output;
[0066] A balance wheel sorting adaptability analysis module, wherein the balance wheel sorting adaptability analysis module is used to use the balance wheel sorting adaptability analyzer to perform a balance wheel sorting adaptability analysis on the balance wheel sorting parameters in combination with the sorting scenario matrix to obtain a balance wheel adaptability matrix;
[0067] A cross-belt sorting adaptation analyzer construction module is used to construct a cross-belt sorting adaptation analyzer, perform cross-belt sorting adaptability analysis on the cross-belt sorting parameters in combination with the sorting scenario matrix, and obtain a cross-belt adaptability matrix.
[0068] Furthermore, the embodiment of the present application also includes:
[0069] A sorting category quantity extraction module, the sorting category quantity extraction module is used to extract the balance wheel sorting category quantity and the cross belt sorting category quantity according to the balance wheel sorting parameter and the cross belt sorting parameter;
[0070] An occupancy saturation calculation module, the occupancy saturation calculation module is used to calculate the occupancy saturation of the sorting category quantity and the balance wheel sorting category quantity and the cross belt sorting category quantity, obtain the balance wheel sorting saturation and the cross belt sorting saturation, and allocate and calculate to obtain a first decision probability distribution;
[0071] An adaptability parameter calculation module, the adaptability parameter calculation module is used to calculate and obtain the balance wheel adaptability parameter and the cross-band adaptability parameter according to the balance wheel adaptability matrix and the cross-band adaptability matrix;
[0072] A decision probability distribution calculation module, the decision probability distribution calculation module is used to allocate and calculate to obtain a second decision probability distribution according to the balance wheel adaptability parameter and the cross belt adaptability parameter;
[0073] The sorting decision probability distribution calculation module is used to calculate and obtain the sorting decision probability distribution according to the first decision probability distribution and the second decision probability distribution;
[0074] A sorting module is used to decide whether to select a pendulum sorting device or a cross-belt sorting device to sort the batch of goods according to the sorting decision probability distribution.
[0075] For a specific embodiment of a scenario-based adaptive sorting decision system, please refer to the embodiment of a scenario-based adaptive sorting decision method described above, which will not be repeated here. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
[0076] In one embodiment, a computer device is provided, which may be a server. The computer device includes a processor, a memory and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store news data and data such as time decay factors. The network interface of the computer device is used to communicate with an external terminal via a network connection. The computer program is executed by the processor to implement a scenario-based adaptive sorting decision method.
[0077] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0078] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application.
Claims
1. A scenario-based adaptive sorting decision method, It is characterized in that The method comprises: Obtaining a batch of goods to be sorted, collecting sorting categories for the batch of goods, and obtaining the number of sorting categories; Collecting images of a plurality of sorted goods in the batch of goods, identifying the goods specifications based on the plurality of goods images, and obtaining a plurality of goods specification parameters; According to the multiple cargo images, recognize the surface friction features of the cargo to obtain multiple surface friction parameters; Collecting the cargo categories of the plurality of sorted cargoes, performing a vulnerability analysis according to the plurality of cargo categories, and obtaining a plurality of vulnerability parameters; According to the plurality of surface friction parameters, the plurality of cargo specification parameters and the plurality of vulnerability parameters, a sorting scenario matrix is constructed, and according to the pre-constructed balance wheel sorting parameters and cross belt sorting parameters of the balance wheel sorting device and the cross belt sorting device, a sorting adaptability analysis is performed to obtain a balance wheel adaptability matrix and a cross belt adaptability matrix; According to the number of sorting categories, a first sorting decision is made to obtain a first decision probability distribution; according to the balance wheel adaptability matrix and the cross belt adaptability matrix, a second sorting decision is made to obtain a second decision probability distribution; in combination with the first decision probability distribution, a sorting decision probability distribution is calculated to obtain a sorting decision probability distribution; and according to the sorting decision probability distribution, a balance wheel sorting device or a cross belt sorting device is selected to sort the batch of goods; According to the plurality of surface friction parameters, the plurality of cargo specification parameters and the plurality of damage parameters, a sorting scenario matrix is constructed, and according to the pre-constructed swing wheel sorting parameters and cross belt sorting parameters of the swing wheel sorting device and the cross belt sorting device, a sorting adaptability analysis is performed, including: Taking the surface friction parameter, the cargo specification parameter and the damage parameter as row attributes and the plurality of sorted cargoes as column attributes, constructing a sorting scenario matrix according to the plurality of surface friction parameters, the plurality of cargo specification parameters and the plurality of damage parameters; Constructing a balance wheel sorting adaptability analyzer, wherein the balance wheel sorting adaptability analyzer includes a balance wheel friction adaptability analysis branch, a balance wheel specification adaptability analysis branch, and a balance wheel fragility adaptability analysis branch, which are respectively constructed by taking a sample balance wheel sorting parameter combined with a sample surface friction parameter, a sample cargo specification parameter, and a sample fragility parameter as inputs, and taking a sample balance wheel friction adaptability parameter, a sample balance wheel specification adaptability parameter, and a sample balance wheel fragility adaptability parameter as outputs; Using the balance wheel sorting adaptability analyzer, the balance wheel sorting parameters are combined with the sorting scenario matrix to perform balance wheel sorting adaptability analysis to obtain a balance wheel adaptability matrix; A cross-belt sorting adaptability analyzer is constructed to perform a cross-belt sorting adaptability analysis on the cross-belt sorting parameters in combination with the sorting scenario matrix to obtain a cross-belt adaptability matrix.
2. The method according to claim 1, It is characterized in that The batch of goods is sorted and collected to obtain the number of sorted categories, including: Extracting multiple sorting directions of the multiple sorted goods according to the sorting identifiers of the multiple sorted goods in the batch of goods; The multiple sorting directions are classified to obtain the number of sorting categories.
3. The method according to claim 1, It is characterized in that Collect images of multiple sorted goods in the batch of goods, identify the goods specifications based on the multiple goods images, and obtain multiple goods specification parameters, including: Collecting images of a plurality of sorted goods in the batch of goods to obtain a plurality of goods images; Obtain a set of sample cargo images, and mark cargo specification and size parameters in the sample cargo images to obtain a set of sample cargo specification parameters; Using the sample cargo image set and the sample cargo specification parameter set, construct a cargo specification identifier; Based on the cargo specification identifier, cargo specification identification is performed on the multiple cargo images to obtain the multiple cargo specification parameters.
4. The method according to claim 1, It is characterized in that According to the multiple cargo images, the cargo surface friction features are identified to obtain multiple surface friction parameters, including: Cropping the multiple cargo images to obtain multiple cargo partial images; Collecting a set of partial images of sample cargo, and marking the cargo surface friction parameters in the partial images of the sample cargo, to obtain a set of sample surface friction parameters; Using the sample cargo local image set and the sample surface friction parameter set, construct a surface friction identifier; The surface friction identifier is used to identify the multiple cargo partial images to obtain the multiple surface friction parameters.
5. The method according to claim 1, It is characterized in that The cargo categories of the plurality of sorted cargoes are collected, and vulnerability analysis is performed according to the plurality of cargo categories to obtain a plurality of vulnerability parameters, including: Collecting the cargo categories of the plurality of sorted cargoes; Obtain a set of sample goods categories, classify them by vulnerability level, and obtain a set of sample vulnerability parameters; Based on the decision tree, the sample cargo category set and the sample vulnerability parameter set are used to construct a vulnerability classifier, and vulnerability analysis is performed on multiple cargo categories to obtain multiple vulnerability parameters.
6. The method according to claim 1, It is characterized in that According to the number of sorting categories, a first sorting decision is made to obtain a first decision probability distribution; according to the balance wheel adaptability matrix and the cross belt adaptability matrix, a second sorting decision is made to obtain a second decision probability distribution; and in combination with the first decision probability distribution, a sorting decision probability is calculated and obtained, including: According to the balance wheel sorting parameters and the cross belt sorting parameters, extract and obtain the balance wheel sorting category quantity and the cross belt sorting category quantity; Calculate the occupancy saturation of the number of sorting categories and the number of balance wheel sorting categories and the number of cross belt sorting categories, obtain the balance wheel sorting saturation and the cross belt sorting saturation, and allocate and calculate to obtain the first decision probability distribution; According to the balance wheel adaptability matrix and the cross-band adaptability matrix, a balance wheel adaptability parameter and a cross-band adaptability parameter are calculated and obtained; According to the balance wheel adaptability parameter and the cross-band adaptability parameter, a second decision probability distribution is obtained by allocation calculation; Calculate and obtain a sorting decision probability distribution according to the first decision probability distribution and the second decision probability distribution; According to the sorting decision probability distribution, a decision is made to select a pendulum wheel sorting device or a cross belt sorting device to sort the batch of goods.
7. A scenario-based adaptive sorting decision system, It is characterized in that The system is used to perform the steps of the method described in any one of claims 1 to 6, and the system comprises: A sorting category collection module, which is used to obtain a batch of goods to be sorted, collect the sorting categories of the batch of goods, and obtain the number of sorting categories; A cargo specification parameter acquisition module, the cargo specification parameter acquisition module is used to collect images of a plurality of sorted cargoes in the batch of cargoes, perform cargo specification recognition based on the plurality of cargo images, and obtain a plurality of cargo specification parameters; a surface friction parameter acquisition module, the surface friction parameter acquisition module being used to identify the surface friction features of the cargo according to the multiple cargo images, and to obtain multiple surface friction parameters; A cargo category collection module, the cargo category collection module is used to collect cargo categories of the plurality of sorted cargoes, perform vulnerability analysis according to the plurality of cargo categories, and obtain a plurality of vulnerability parameters; A sorting adaptability analysis module, wherein the sorting adaptability analysis module is used to construct a sorting scenario matrix according to the plurality of surface friction parameters, the plurality of cargo specification parameters and the plurality of vulnerability parameters, and to perform sorting adaptability analysis according to the pre-constructed balance wheel sorting parameters and cross belt sorting parameters of the balance wheel sorting device and the cross belt sorting device, so as to obtain a balance wheel adaptability matrix and a cross belt adaptability matrix; A sorting decision probability calculation module, the sorting decision probability calculation module is used to make a first sorting decision according to the number of sorting categories to obtain a first decision probability distribution, make a second sorting decision according to the balance wheel adaptability matrix and the cross belt adaptability matrix to obtain a second decision probability distribution, calculate and obtain a sorting decision probability distribution in combination with the first decision probability distribution, and select a balance wheel sorting device or a cross belt sorting device to sort the batch of goods according to the sorting decision probability distribution; According to the plurality of surface friction parameters, the plurality of cargo specification parameters and the plurality of damage parameters, a sorting scenario matrix is constructed, and according to the pre-constructed swing wheel sorting parameters and cross belt sorting parameters of the swing wheel sorting device and the cross belt sorting device, a sorting adaptability analysis is performed, including: Taking the surface friction parameter, the cargo specification parameter and the damage parameter as row attributes and the plurality of sorted cargoes as column attributes, constructing a sorting scenario matrix according to the plurality of surface friction parameters, the plurality of cargo specification parameters and the plurality of damage parameters; Constructing a balance wheel sorting adaptability analyzer, wherein the balance wheel sorting adaptability analyzer includes a balance wheel friction adaptability analysis branch, a balance wheel specification adaptability analysis branch, and a balance wheel fragility adaptability analysis branch, which are respectively constructed by taking a sample balance wheel sorting parameter combined with a sample surface friction parameter, a sample cargo specification parameter, and a sample fragility parameter as inputs, and taking a sample balance wheel friction adaptability parameter, a sample balance wheel specification adaptability parameter, and a sample balance wheel fragility adaptability parameter as outputs; Using the balance wheel sorting adaptability analyzer, the balance wheel sorting parameters are combined with the sorting scenario matrix to perform balance wheel sorting adaptability analysis to obtain a balance wheel adaptability matrix; A cross-belt sorting adaptability analyzer is constructed to perform a cross-belt sorting adaptability analysis on the cross-belt sorting parameters in combination with the sorting scenario matrix to obtain a cross-belt adaptability matrix.
8. A computer device comprising a memory and a processor, wherein the memory stores a computer program. It is characterized in that When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.
Citation Information
Patent Citations
Order distribution method and device, electronic equipment and storage medium
CN109711940A
Order processing and shelf assignment method for intelligent warehouse
CN111815040A
Sorting method and device, equipment and storage medium
CN112101833A
Method, System, and Apparatus for Categorizing and Presenting Risk Based Analytical Results
US20100198632A1
System and method for real time assessment of cargo handling
US20150254600A1
