Equipment scheduling method and system for express delivery sorting
By setting up an image acquisition module in the express sorting system and establishing a device tag library, identifying the express type and matching the equipment, the problem of inflexible equipment scheduling in traditional systems is solved, and sorting efficiency and accuracy are improved.
Patent Information
- Application Number
- CN202411406718.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-10
- Publication Date
- 2024-12-20
- Estimated Expiration
- 2044-10-10
AI Technical Summary
Traditional express sorting systems cannot intelligently schedule corresponding equipment for efficient and accurate sorting according to the express type, and lack real-time equipment status information and flexibility, resulting in the impact of sorting efficiency and accuracy.
By setting up an image acquisition module at the initial end of the express sorting line, image acquisition and feature convolution are performed, express type is identified, and sorting equipment label library is established to match appropriate sorting equipment to realize equipment scheduling.
It realizes automatic scheduling of corresponding equipment for sorting according to express delivery type, improves sorting efficiency and accuracy, and solves the problem of inflexible equipment scheduling in traditional systems.
Smart Images
Figure CN118918574B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of express sorting, and in particular to an equipment scheduling method and system for express sorting. Background Art
[0002] In the modern logistics industry, sorting equipment plays a vital role. They are responsible for quickly and accurately classifying and distributing a large number of express deliveries according to specific indicators and rules. With the rapid development of e-commerce, the number of express deliveries has exploded, which has put forward higher requirements for the efficiency and accuracy of the sorting system. In order to meet this demand, the automation and intelligence of express sorting systems have become an inevitable trend in the development of the industry. In traditional logistics sorting technology, a significant defect is that it is difficult to intelligently dispatch corresponding equipment according to the type of express delivery for efficient and accurate sorting. Traditional sorting systems cannot automatically identify the type and attributes of express delivery, such as weight, size, destination, etc., resulting in the inability to intelligently dispatch the most suitable sorting equipment for processing based on this information. Secondly, traditional sorting systems lack real-time updated equipment status information and cannot be dynamically dispatched according to the current status of the equipment (such as idle, busy, faulty, etc.). In addition, traditional sorting systems often use fixed sorting processes and rules, lacking flexibility and adaptability. When the type of express delivery or sorting requirements change, the system cannot quickly adjust and optimize the sorting scheme, resulting in serious impact on sorting efficiency and accuracy. Summary of the invention
[0003] The embodiments of the present application provide a method and system for scheduling equipment for express delivery sorting, which solves the technical problem in the prior art that it is difficult to intelligently schedule corresponding equipment according to the type of express delivery for efficient and accurate sorting.
[0004] In view of the above problems, an embodiment of the present application provides an equipment scheduling method and system for express delivery sorting.
[0005] A first aspect of an embodiment of the present application provides a device scheduling method for express sorting, the method comprising:
[0006] An image acquisition module is provided, and image acquisition is performed through the image acquisition module to obtain an image data set, wherein the image acquisition module is provided at the initial end of the express sorting line; feature convolution is performed on the image data set to identify a first type of express indicator and a second type of express indicator, wherein the first type of express indicator is box express, and the second type of express indicator is bag express; a sorting device label library is established, and the sorting device label library is obtained by identifying the gripper features of the sorting device; the first type of express indicator and the second type of express indicator are input into a sorting device dispatching center, and the sorting device dispatching center is connected to the sorting device label library; and the equipment dispatching results of the first type of express indicator and the equipment dispatching results of the second type of express indicator are obtained according to the sorting device label library.
[0007] A second aspect of an embodiment of the present application provides a device scheduling system for express sorting, the system comprising:
[0008] An image acquisition module is set up, and image acquisition is performed through the image acquisition module to obtain an image data set, wherein the image acquisition module is set at the initial end of the express sorting line; an identification module, the identification module is used to perform feature convolution on the image data set, and identify a first type of express indicator and a second type of express indicator, wherein the first type of express indicator is box express, and the second type of express indicator is bag express; a label library establishment module, the label library establishment module is used to establish a sorting equipment label library, and the sorting equipment label library is obtained by identifying the gripper features of the sorting equipment; an indicator input module, the indicator input module is used to input the first type of express indicator and the second type of express indicator into a sorting equipment scheduling center, and the sorting equipment scheduling center is connected to the sorting equipment label library; a scheduling result acquisition module, the scheduling result acquisition module is used to obtain the equipment scheduling results of the first type of express indicators and the equipment scheduling results of the second type of express indicators according to the sorting equipment label library.
[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0010] First, an image acquisition module is set up, and image acquisition is performed through the image acquisition module to obtain an image data set, wherein the image acquisition module is set at the initial end of the express sorting line. Next, feature convolution is performed on the image data set to identify the first type of express indicators and the second type of express indicators, wherein the first type of express indicators are box express, and the second type of express indicators are bag express. Then, a sorting device label library is established, and the sorting device label library is obtained by identifying the gripper features of the sorting device. The first type of express indicators and the second type of express indicators are input into the sorting device dispatching center, and the sorting device dispatching center is connected to the sorting device label library. Finally, the equipment dispatching results of the first type of express indicators and the equipment dispatching results of the second type of express indicators are obtained according to the sorting device label library. The technical problem that it is difficult to intelligently dispatch corresponding equipment according to the express type for efficient and accurate sorting in the prior art is solved, and the technical effect of automatically dispatching corresponding equipment for sorting according to the express type and improving the sorting efficiency is achieved. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] 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.
[0012] Figure 1 A schematic diagram of a flow chart of a device scheduling method for express delivery sorting provided in an embodiment of the present application;
[0013] Figure 2 A schematic diagram of the structure of an equipment scheduling system for express sorting provided in an embodiment of the present application.
[0014] Explanation of reference numerals: image acquisition module 11 , recognition module 12 , label library establishment module 13 , indicator input module 14 , scheduling result acquisition module 15 . DETAILED DESCRIPTION
[0015] The embodiments of the present application solve the technical problem in the prior art that it is difficult to intelligently schedule corresponding equipment according to the type of express delivery for efficient and accurate sorting by providing a device scheduling method and system for express delivery sorting.
[0016] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.
[0017] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or are inherent to these processes, methods, products or devices.
[0018] Embodiment 1
[0019] like Figure 1 As shown, the embodiment of the present application provides a device scheduling method for express sorting, wherein the method includes:
[0020] An image acquisition module is set up, and image acquisition is performed through the image acquisition module to obtain an image data set, wherein the image acquisition module is set at the initial end of the express sorting line.
[0021] In the express sorting process, efficiently and accurately identifying and sorting different types of express parcels is the key to improving the overall logistics efficiency. To this end, an image acquisition module is set up at the initial end of the express sorting line. The main responsibility of the image acquisition module is to perform real-time image acquisition to obtain an image data set containing detailed information of the express parcel. When the express parcel passes through the initial end of the sorting line, the image acquisition module will start immediately and perform continuous image acquisition. After preliminary processing, the collected image data will be saved as an image data set for subsequent express parcel type identification and sorting equipment scheduling. By setting up an image acquisition module at the initial end of the express sorting line, express parcels can be identified in real time and automatically, providing accurate and reliable data support for subsequent sorting operations.
[0022] Feature convolution is performed on the image data set to identify first-category express indicators and second-category express indicators, wherein the first-category express indicators are box express and the second-category express indicators are bag express.
[0023] Feature convolution is performed on the image dataset to identify the categories of express parcels, including the first category and the second category. Among them, the first category is box express, and the second category is bag express. Feature convolution is a commonly used image processing technology in the field of deep learning. It simulates the operating mechanism of the human brain's visual system to extract multi-level features of the image. Specifically, the image dataset is input into a pre-trained deep learning model, which contains multiple convolutional layers, pooling layers, and fully connected layers. When the image data passes through the convolution layer, a series of feature maps are generated, which describe the feature information of the image at different levels. Subsequently, the feature map is reduced in dimension and feature selected through the pooling layer to reduce the amount of calculation and improve the robustness of the model. Finally, the fully connected layer integrates the extracted feature information and outputs the final classification result. Through feature convolution processing, the model can accurately identify box express and bag express in the image dataset.
[0024] A sorting device label library is established, and the sorting device label library is obtained by identifying the gripper features of the sorting device.
[0025] In the express sorting system, in order to improve the sorting efficiency and accuracy, it is necessary to establish a corresponding label library for each sorting device. The sorting device label library is obtained by identifying the gripper features of the sorting device, aiming to ensure that the system can accurately match the express parcel with the appropriate sorting device. Specifically, the gripper features of the sorting device are identified and analyzed in detail, including key information such as the shape, size, material, and gripping method of the gripper. These features will directly affect what type and size of express parcels the device can handle. These identified gripper features are encoded and stored to form a sorting device label library. Each label corresponds to a specific sorting device and contains detailed information and feature descriptions of the device. In this way, when the system receives image data of an express parcel, it can find the most suitable sorting device in the label library by feature matching.
[0026] Furthermore, a sorting equipment label library is established, and the method includes:
[0027] Train a gripper feature recognition model, use the gripper feature recognition model to collect data on the grippers of each sorting device, and obtain the gripper opening and closing degree, gripper tightness and gripper compatibility; classify the labels of each sorting device based on the gripper opening and closing degree, gripper tightness and gripper compatibility, and output the label results of each sorting device, wherein each sorting setting includes at least one label; and establish a sorting device label library according to the label results of each sorting device.
[0028] The gripper feature recognition model can collect data and classify labels for the grippers of each sorting device. Using the gripper feature recognition model, data is collected for the grippers of each sorting device to obtain the data of each gripper, including gripper opening and closing, gripper tightness and gripper compatibility. The gripper opening and closing refers to the maximum and minimum distance that the gripper finger (or jaw) can open; gripper tightness refers to the gripping force or tightness that the gripper can provide when gripping an item; gripper compatibility means that the gripper needs to work with other parts of the sorting device (such as robot arms, conveyor belts, etc.), and the gripper requirements for box express and bag express are different. Using the characteristics of opening and closing, tightness and compatibility as the basis for classification, one or more labels are assigned to each sorting device. These labels describe the type of express that the device can handle, the size range, and special functions. According to the label results of each sorting device, a sorting device label library is established. The sorting device label library is used to store and manage the label information of all sorting devices. Each device has a unique identifier and a corresponding label set in the label library. By querying the label library, the system can quickly and accurately obtain the detailed information and applicable scope of each device. In actual applications, when the system receives a courier package to be sorted, it can search for matching sorting equipment in the label library based on the package's size, shape, material and other feature information. Then, the system will dispatch these devices to automatically sort the package.
[0029] Furthermore, the gripper feature recognition model is trained, and the method includes:
[0030] Acquire a sorting device sample, collect data on the sorting device sample with the gripper opening and closing degree, gripper tightness and gripper compatibility as identification features, and generate a feature sample data set; set a random cutoff point, and the random cutoff point is used to determine the label category to which the sample belongs; predict the feature sample data set according to the random cutoff point, and output the predicted label of the sorting device sample; compare the predicted label of the sorting device sample with the real label of the sorting device sample, and output the prediction accuracy; until the prediction accuracy reaches a preset threshold, the model converges and the gripper feature recognition model is output.
[0031] Construct a training gripper feature recognition model. Specifically, collect sorting equipment samples, which cover sorting equipment of different models, sizes and functions. Take gripper opening and closing degree, gripper tightness and gripper compatibility as identification features, and collect data for sorting equipment samples. Use sensors, cameras or other measuring devices to obtain the precise values of these features and generate feature sample data sets. Set random cutoff points to determine the label category to which the sample belongs. Use the feature sample data set and the set random cutoff points to train the model, predict the sorting equipment samples, and output the predicted labels. Compare the predicted labels of the sorting equipment samples with the true labels of the sorting equipment samples, and calculate the prediction accuracy. The prediction accuracy is the ratio of the number of samples predicted correctly by the model to the total number of samples. Evaluate the performance of the model based on the prediction accuracy. If the prediction accuracy is lower than the preset threshold, it means that the model needs to be further optimized. At this time, use optimization algorithms such as gradient descent to update the parameters of the model so that the model can better fit the data. Repeat the above prediction, comparison, evaluation and update process until the prediction accuracy of the model reaches the preset threshold or the maximum number of iterations is reached. When the prediction accuracy of the model reaches the preset threshold, the model is considered to have converged and has good performance. At this time, the gripper feature recognition model is output.
[0032] Furthermore, the method further comprises:
[0033] The random truncation points include a first random truncation point and a second random truncation point; predicting the sorting device sample, and outputting the predicted probability of the sorting device sample belonging to the first category; making a judgment based on the first random truncation point and the second random truncation point and the predicted probability, and outputting the judgment result; and outputting the predicted label of the sorting device sample according to the judgment result.
[0034] The random cutoff point (also called the threshold) is a standard used to convert the predicted probability of the model into a specific label. The random cutoff point includes the first random cutoff point and the second random cutoff point, and the random cutoff point is used to determine the label category to which the sample belongs. The feature data of the sorting device sample is input into the trained gripper feature recognition model, and the model outputs the predicted probability of the sorting device sample belonging to the first category. Based on the comparison between the first random cutoff point and the second random cutoff point and the predicted probability, a judgment is made and the judgment result is output. According to the judgment result, the predicted label of the sorting device sample is finally determined. It should be noted that the setting of the random cutoff point has a great influence on the accuracy of the prediction result. In practical applications, it is necessary to select a suitable cutoff point value according to the specific situation and verify its effectiveness through experiments. At the same time, with the continuous accumulation of data and the continuous optimization of the model, the value of the cutoff point needs to be regularly updated and adjusted.
[0035] Further, judging based on the first random truncation point, the second random truncation point and the predicted probability includes:
[0036] When the predicted probability of the sorting device is less than the probability of the first random truncation point, it is marked as the second category label; when the predicted probability of the sorting device is greater than the probability of the second random truncation point, it is marked as the first category label; when the predicted probability of the sorting device is greater than the probability of the first random truncation point and less than the probability of the second random truncation point, it is marked as the first category label / the second category label at the same time.
[0037] When the predicted probability of the sorting device is less than the probability of the first random cutoff point, the sorting device is identified as the second category label, and the second category label is a device that only sorts bags. When the predicted probability of the sorting device is greater than the probability of the second random cutoff point, the sorting device is identified as the first category label, and the first category label is a device that only sorts boxes. When the predicted probability of the sorting device is greater than the probability of the first random cutoff point and less than the probability of the second random cutoff point, the sorting device is identified as both the first category label and the second category label. The first category label / second category label indicates that this type of equipment can sort bags and boxes at the same time.
[0038] Further, after outputting the prediction accuracy, the method includes:
[0039] When the prediction accuracy rate does not reach a preset threshold, a gradient descent step is set; minimizing the difference between the prediction accuracy rate and the preset threshold rate is taken as an optimization target, the first random truncation point and the second random truncation point are used as optimization variables, gradient optimization is performed based on the gradient descent step, and the optimized first random truncation point and the second random truncation point are output.
[0040] In the process of training the gripper feature recognition model, if the prediction accuracy does not reach the preset threshold, the model parameters can be adjusted based on the gradient descent algorithm to optimize the performance of the model. Specifically, a suitable gradient descent step size is set to determine the size of the variable to be updated in each iteration, with minimizing the difference between the prediction accuracy and the preset threshold as the optimization goal, and the first random truncation point and the second random truncation point are used as optimization variables. The gradients of the first random truncation point and the second random truncation point are calculated. In each iteration, the first random truncation point and the second random truncation point are updated in the direction of the gradient descent according to the calculated gradient and the set step size. After multiple iterations, the first random truncation point and the second random truncation point after optimization are obtained.
[0041] The first category express indicators and the second category express indicators are input into a sorting device dispatching center, and the sorting device dispatching center is connected to the sorting device label library.
[0042] The sorting equipment dispatch center is connected to the sorting equipment label library to optimize and dispatch the operation of the sorting equipment according to various indicators and labels. The first-class express indicators and the second-class express indicators are input into the sorting equipment dispatch center to optimize and dispatch the operation of the sorting equipment according to the input express indicators and the equipment information in the sorting equipment label library.
[0043] The equipment scheduling results of the first category express indicators and the equipment scheduling results of the second category express indicators are obtained according to the sorting equipment label library.
[0044] The equipment scheduling results of the first-class express indicators and the second-class express indicators are obtained according to the sorting equipment label library to ensure the smooth progress and efficient completion of the sorting task.
[0045] Furthermore, the method for obtaining the equipment scheduling result of the first type of express delivery indicator and the equipment scheduling result of the second type of express delivery indicator according to the sorting equipment label library includes:
[0046] According to the first category express indicators, tasks of sorting devices with first category labels are configured in the sorting device label library; according to the second category express indicators, tasks of sorting devices with second category labels are configured in the sorting device label library.
[0047] When the system receives express information with the first category of express indicators, it will query the sorting device label library, find the sorting device with the first category label, and assign the corresponding task. For the second category of express indicators, a similar method will be used to query the sorting device label library to find the sorting device with the second category label, and configure the corresponding sorting task for the selected device. Through the above steps, the system can configure the sorting device tasks with the corresponding category labels in the sorting device label library according to different types of express indicators, thereby achieving efficient and accurate sorting operations.
[0048] In summary, the embodiments of the present application have at least the following technical effects:
[0049] First, an image acquisition module is set up, and image acquisition is performed through the image acquisition module to obtain an image data set, wherein the image acquisition module is set at the initial end of the express sorting line. Next, feature convolution is performed on the image data set to identify the first type of express indicators and the second type of express indicators, wherein the first type of express indicators are box express, and the second type of express indicators are bag express. Then, a sorting device label library is established, and the sorting device label library is obtained by identifying the gripper features of the sorting device. The first type of express indicators and the second type of express indicators are input into the sorting device dispatching center, and the sorting device dispatching center is connected to the sorting device label library. Finally, the equipment dispatching results of the first type of express indicators and the equipment dispatching results of the second type of express indicators are obtained according to the sorting device label library. The technical problem that it is difficult to intelligently dispatch corresponding equipment according to the express type for efficient and accurate sorting in the prior art is solved, and the technical effect of automatically dispatching corresponding equipment for sorting according to the express type and improving the sorting efficiency is achieved.
[0050] Embodiment 2
[0051] Based on the same inventive concept as the device scheduling method for express sorting in the aforementioned embodiment, Figure 2 As shown, the present application provides a device scheduling system for express sorting, and the system and method embodiments in the present application are based on the same inventive concept. The system includes:
[0052] An image acquisition module 11 is provided, and image acquisition is performed through the image acquisition module 11 to obtain an image data set, wherein the image acquisition module is provided at the initial end of the express sorting line; an identification module 12, and the identification module 12 is used to perform feature convolution on the image data set to identify a first type of express indicator and a second type of express indicator, wherein the first type of express indicator is box express, and the second type of express indicator is bag express; a label library establishment module 13, and the label library establishment module 13 is used to establish a sorting device label library, and the sorting device label library is obtained by identifying the gripper features of the sorting device; an indicator input module 14, and the indicator input module 14 is used to input the first type of express indicator and the second type of express indicator into a sorting device dispatching center, and the sorting device dispatching center is connected to the sorting device label library; a scheduling result acquisition module 15, and the scheduling result acquisition module 15 is used to obtain the equipment scheduling result of the first type of express indicator and the equipment scheduling result of the second type of express indicator according to the sorting device label library.
[0053] Furthermore, the label library establishment module 13 is used to execute the following method:
[0054] Train a gripper feature recognition model, use the gripper feature recognition model to collect data on the grippers of each sorting device, and obtain the gripper opening and closing degree, gripper tightness and gripper compatibility; classify the labels of each sorting device based on the gripper opening and closing degree, gripper tightness and gripper compatibility, and output the label results of each sorting device, wherein each sorting setting includes at least one label; and establish a sorting device label library according to the label results of each sorting device.
[0055] Furthermore, the label library establishment module 13 is used to execute the following method:
[0056] Acquire a sorting device sample, collect data on the sorting device sample with the gripper opening and closing degree, gripper tightness and gripper compatibility as identification features, and generate a feature sample data set; set a random cutoff point, and the random cutoff point is used to determine the label category to which the sample belongs; predict the feature sample data set according to the random cutoff point, and output the predicted label of the sorting device sample; compare the predicted label of the sorting device sample with the real label of the sorting device sample, and output the prediction accuracy; until the prediction accuracy reaches a preset threshold, the model converges and the gripper feature recognition model is output.
[0057] Furthermore, the label library establishment module 13 is used to execute the following method:
[0058] Predict the sorting device sample and output the predicted probability of the sorting device sample belonging to the first category; make a judgment based on the first random truncation point and the second random truncation point and the predicted probability, and output the judgment result; according to the judgment result, output the predicted label of the sorting device sample.
[0059] Furthermore, the label library establishment module 13 is used to execute the following method:
[0060] When the predicted probability of the sorting device is less than the probability of the first random truncation point, it is marked as the second category label; when the predicted probability of the sorting device is greater than the probability of the second random truncation point, it is marked as the first category label; when the predicted probability of the sorting device is greater than the probability of the first random truncation point and less than the probability of the second random truncation point, it is marked as the first category label / the second category label at the same time.
[0061] Furthermore, the label library establishment module 13 is used to execute the following method:
[0062] When the prediction accuracy rate does not reach a preset threshold, a gradient descent step is set; minimizing the difference between the prediction accuracy rate and the preset threshold rate is taken as an optimization target, the first random truncation point and the second random truncation point are used as optimization variables, gradient optimization is performed based on the gradient descent step, and the optimized first random truncation point and the second random truncation point are output.
[0063] Furthermore, the scheduling result acquisition module 15 is used to execute the following method:
[0064] According to the first category express indicators, tasks of sorting devices with first category labels are configured in the sorting device label library; according to the second category express indicators, tasks of sorting devices with second category labels are configured in the sorting device label library.
[0065] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0066] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.
[0067] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.
Claims
1. A device scheduling method for express sorting, characterized in that: The method comprises: An image acquisition module is provided, and image acquisition is performed through the image acquisition module to obtain an image data set, wherein the image acquisition module is provided at the initial end of the express sorting line; Perform feature convolution on the image data set to identify a first type of express delivery indicator and a second type of express delivery indicator, wherein the first type of express delivery indicator is box express delivery, and the second type of express delivery indicator is bag express delivery; Establishing a sorting equipment label library, wherein the sorting equipment label library is obtained by identifying the gripper features of the sorting equipment; Inputting the first category express indicators and the second category express indicators into a sorting device dispatching center, wherein the sorting device dispatching center is connected to the sorting device label library; Acquire the equipment scheduling result of the first type of express indicators and the equipment scheduling result of the second type of express indicators according to the sorting equipment label library; The method for establishing a sorting equipment label library includes: Training a gripper feature recognition model, and using the gripper feature recognition model to collect data on the grippers of various sorting devices to obtain gripper opening and closing degrees, gripper tightness, and gripper compatibility; Classify the various sorting devices by labeling them based on the gripper opening and closing degree, gripper tightness and gripper compatibility, and output the label results of the various sorting devices, wherein each sorting setting includes at least one label; According to the labeling results of each sorting device, a sorting device label library is established.
2. The method according to claim 1, characterized in that Training the gripper feature recognition model includes: Acquire a sorting device sample, collect data on the sorting device sample using the gripper opening and closing degree, gripper tightness and gripper compatibility as identification features, and generate a feature sample data set; Set a random cutoff point, which is used to determine the label category to which the sample belongs; Predicting the feature sample data set according to the random cutoff point, and outputting the predicted label of the sorting device sample; Compare the predicted label of the sorting device sample with the real label of the sorting device sample, and output the prediction accuracy; The model converges until the prediction accuracy reaches a preset threshold, and the gripper feature recognition model is output.
3. The method according to claim 2, characterized in that The random truncation points include a first random truncation point and a second random truncation point; Predicting the sorting device sample, and outputting the predicted probability of the sorting device sample belonging to the first category; Perform judgment based on the first random truncation point, the second random truncation point and the predicted probability, and output a judgment result; According to the judgment result, the predicted label of the sorting device sample is output.
4. The method according to claim 3, characterized in that The determination based on the first random truncation point, the second random truncation point and the predicted probability includes: When the predicted probability of the sorting device is less than the probability of the first random cutoff point, it is marked as a second category label; When the predicted probability of the sorting device is greater than the probability of the second random cutoff point, it is marked as a first category label; When the predicted probability of the sorting device is greater than the probability of the first random truncation point and less than the probability of the second random truncation point, it is simultaneously identified as the first category label / the second category label.
5. The method according to claim 3, characterized in that After outputting the prediction accuracy, the method includes: When the prediction accuracy does not reach a preset threshold, setting a gradient descent step size; Taking minimizing the difference between the prediction accuracy and the preset threshold as the optimization goal, taking the first random truncation point and the second random truncation point as optimization variables, performing gradient optimization based on the gradient descent step size, and outputting the optimized first random truncation point and the second random truncation point.
6. The method according to claim 1, characterized in that The method of obtaining the equipment scheduling result of the first type of express indicators and the equipment scheduling result of the second type of express indicators according to the sorting equipment label library includes: According to the first category of express delivery indicators, the task of configuring the sorting device with the first category of labels in the sorting device label library; According to the second category express indicators, the task of sorting equipment with second category labels is configured in the sorting equipment label library.
7. Equipment dispatching system for express sorting, characterized in that: The system is used to implement the equipment scheduling method for express sorting according to any one of claims 1 to 6, and comprises: An image acquisition module is provided, and image acquisition is performed through the image acquisition module to obtain an image data set, wherein the image acquisition module is provided at the initial end of the express sorting line; A recognition module, the recognition module is used to perform feature convolution on the image data set to identify a first type of express delivery indicator and a second type of express delivery indicator, wherein the first type of express delivery indicator is box express delivery, and the second type of express delivery indicator is bag express delivery; A label library establishment module, which is used to establish a label library for sorting equipment. The label library for sorting equipment is obtained by identifying the gripper features of the sorting equipment; An indicator input module, the indicator input module is used to input the first category express indicators and the second category express indicators into a sorting device dispatching center, the sorting device dispatching center is connected to the sorting device label library; A scheduling result acquisition module is used to obtain the equipment scheduling results of the first type of express indicators and the equipment scheduling results of the second type of express indicators according to the sorting device label library.