A package mis-sorting early warning and anti-jamming control method and system

Abstract

The embodiment of the specification discloses a package mis-sorting early warning and anti-blocking control method and system, and relates to the technical fields of logistics management and automatic control. The method comprises the following steps: when a package enters a sorting loop, obtaining classification information, posture information and current position information of the package; determining a target bin according to the classification information, and determining target position information according to the target bin; predicting a time range in which the package triggers a bin drop sensor of the target bin according to the posture information, the current position information and the target position information; after sending a bin drop command for the package to a sorting device, monitoring bin drop sensor signals of the target bin and its adjacent bins in real time; if it is monitored that the target bin does not trigger its bin drop sensor signal within the predicted time range, and no predetermined dropped package is found in the adjacent bins within or near the time range, but the bin drop sensor signals of the adjacent bins are triggered, it is determined that the package has a mis-sorting phenomenon, and a mis-sorting early warning is triggered.

Description

Technical Field

[0001] This application relates to the fields of logistics management and automatic control technology, and more specifically, to a method and system for early warning and prevention of package misclassification. Background Technology

[0002] With the rapid development of the e-commerce industry, the number of packages handled by logistics sorting centers is increasing day by day, and therefore, the requirements for sorting efficiency and accuracy are also getting higher and higher.

[0003] In automated sorting systems, package missorting (such as "floating" or "random sorting") and compartment blockage are the main problems affecting sorting efficiency. While existing automated sorting systems can improve efficiency to some extent, they still struggle to completely eliminate package missorting and compartment blockage when handling massive volumes of packages. Once these problems occur during sorting, they not only lead to a significant drop in efficiency but can also trigger a series of chain reactions, including package damage (caused by blockage), delayed delivery (caused by missorting), and customer complaints, resulting in substantial economic losses and reputational damage for logistics companies. Therefore, developing a control method and system that can effectively warn of package missorting and prevent compartment blockage is of paramount importance. Summary of the Invention

[0004] To achieve the above objectives, one aspect of the embodiments of this specification provides a method for package misclassification warning and anti-blocking control, the method comprising:

[0005] When a package enters the sorting loop, its classification information, posture information, and current location information are obtained.

[0006] The target compartment corresponding to the package is determined based on the classification information, and the target location information is determined based on the location of the target compartment.

[0007] Based on the attitude information, current position information, and target position information, predict the time range within which the package triggers the drop sensor of the target compartment.

[0008] After sending a sorting command for the package to the sorting equipment, the sorting sensor signals of the target compartment and its adjacent compartments are monitored in real time.

[0009] If the target compartment does not trigger its parcel placement sensor signal within the predicted time range, and the adjacent compartment does not have a pre-determined parcel in or near the time range, but triggers its parcel placement sensor signal in or near the time range, then it is determined that the parcel has been missorted, and a missorting warning is triggered.

[0010] In some embodiments, obtaining the package's classification information, posture information, and current location information includes:

[0011] Image of the package is acquired using a correction camera;

[0012] Based on the package image, the barcode, size, packaging type, and destination information of the package are identified. Then, based on at least one of the barcode, size, packaging type, destination information, and the RFID tag information of the package, the classification information of the package is determined.

[0013] Based on the package image, the relative position information of the package relative to the sorting loading table is identified, and then the posture information corresponding to the package is obtained based on the relative position information;

[0014] Based on the distribution location information of the sorting loading platform in the sorting loop, the current location information of the package is determined.

[0015] In some embodiments, predicting the time range within which the package triggers the landing sensor of the target compartment based on the attitude information, current location information, and target location information includes:

[0016] Based on the current location information, the target location information, and the operating speed information of the sorting loop, predict the standard reference time range required for the package to reach the target compartment;

[0017] Based on the attitude information, the standard reference time range for the package to arrive at the target compartment is corrected to obtain the time range within which the package triggers the compartment sensor; wherein,

[0018] The step of correcting the standard reference time range for the arrival of the package at the target compartment includes: taking the midpoint of the standard reference time range as the center, and then determining the correction magnitude based on the error time calculated based on the attitude information, thereby determining a time interval as the time range for the package to trigger the landing sensor of the target compartment.

[0019] In some embodiments, each compartment is provided with a buffer area connected to it, the buffer area being used to temporarily store the latest parcel that has been placed in the compartment, and the method further includes: when a missorted parcel is detected flowing into any compartment, causing the latest parcel that has been placed in the corresponding buffer area to flow back into the sorting loop.

[0020] In some embodiments, the method further includes:

[0021] Real-time monitoring of the signal status of the sensor at each grid opening;

[0022] When the sensor signal of a certain grid is continuously triggered and exceeds the preset first time threshold, it is determined that the grid is blocked and a grid blockage alarm is triggered.

[0023] In some embodiments, the method further includes: when a misclassified package is detected flowing into any compartment or a blockage occurs, delaying the placement time of the relevant package to be placed in the compartment.

[0024] In some embodiments, the method further includes:

[0025] Acquire historical sorting data and construct a training sample set, wherein the historical sorting data includes the classification information, posture information and corresponding sorting result information of historical packages;

[0026] A misclassification risk prediction model is trained based on the training sample set. The misclassification risk prediction model is configured to output the corresponding misclassification risk assessment result based on the classification information and posture information of the packages entering the sorting loop.

[0027] For packages entering the sorting loop, the trained misclassification risk prediction model is used to determine the corresponding misclassification risk assessment result.

[0028] Based on the missorting risk assessment results, a predictive handling strategy is implemented for high-risk packages; wherein, the predictive handling strategy includes any one of posture correction, manual intervention, and re-sorting.

[0029] In some embodiments, the misclassification risk prediction model is trained in the following manner:

[0030] Using the classification information and posture information corresponding to the historical packages as input features, and the misclassification phenomenon recognition results reflected by the sorting result information corresponding to the historical packages as labels, a training sample set is constructed;

[0031] The initial machine learning model is trained using the training sample set, and the parameters of the initial machine learning model are continuously adjusted through an optimization algorithm so that the error between the misclassification risk assessment result output by the model and the label gradually decreases; wherein, the initial machine learning model includes any one of a neural network model, a support vector machine model, or a random forest model, and the optimization algorithm includes the gradient descent algorithm;

[0032] When the error between the model's output misclassification risk assessment result and the label is less than a preset threshold, or when the training reaches the preset maximum number of iterations, training stops, and the trained misclassification risk prediction model is obtained.

[0033] Another aspect of the embodiments of this specification provides a package missorting warning and anti-blocking control system, the system comprising:

[0034] The information acquisition module is used to acquire the package's classification information, posture information, and current location information when the package enters the sorting loop.

[0035] The target compartment determination module is used to determine the target compartment corresponding to the package based on the classification information, and to determine the target location information based on the location of the target compartment;

[0036] The time prediction module is used to predict the time range within which the package triggers the cell sensor of the target compartment based on the attitude information, current position information, and target position information.

[0037] The signal monitoring module is used to monitor the drop-off sensor signals of the target compartment and its adjacent compartments in real time after sending a drop-off command to the sorting equipment for the package.

[0038] The misclassification determination module is used to determine that the package has been misclassified when it is detected that the target compartment has not triggered its parcel sensor signal within the predicted time range, and the adjacent compartment has no predetermined parcels within or near the time range, but has triggered its parcel sensor signal within or near the time range. At the same time, it triggers a misclassification warning.

[0039] In some embodiments, the system further includes a grid blocking determination module, the grid blocking determination module being used for:

[0040] Real-time monitoring of the signal status of the sensor at each grid opening;

[0041] When the sensor signal of a certain grid is continuously triggered and exceeds the preset first time threshold, it is determined that the grid is blocked and a grid blockage alarm is triggered.

[0042] The beneficial effects of the package missorting warning and anti-blocking control method and system provided in the embodiments of this specification include at least the following: by predicting the time range of the package triggering the landing sensor of the target compartment based on the package's classification information, posture information and current location information, and then after sending the landing command for the package to the sorting equipment, the landing sensor signals of the target compartment and its adjacent compartments are monitored in real time, and finally the triggering status of the landing sensor signals of the target compartment and its adjacent compartments within the time range is used to determine whether the package is missorted. This method can realize the monitoring of package missorting at low cost, and greatly reduces the implementation cost of package missorting and blockage monitoring.

[0043] Additional features will be set forth in part in the description which follows. They will become apparent to those skilled in the art upon consulting the following description and the accompanying drawings, or may be learned by the generation or operation of examples. The features of this specification can be realized and obtained through practice or by using various aspects of the methods, tools, and combinations illustrated in the following detailed examples. Attached Figure Description

[0044] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0045] Figure 1 This is an exemplary flowchart of a package misclassification warning and anti-blocking control method according to some embodiments of this specification;

[0046] Figure 2 This is an exemplary flowchart of a package misclassification warning and anti-blocking control method according to other embodiments of this specification;

[0047] Figure 3 This is an exemplary training flowchart of a misclassification risk prediction model according to some embodiments of this specification;

[0048] Figure 4 This is an exemplary block diagram of a package missorting warning and anti-blocking control system according to some embodiments of this specification. Detailed Implementation

[0049] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0050] It should be understood that the terms "system," "device," "unit," and / or "module" as used in this specification are a method of distinguishing different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0051] As indicated in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0052] Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0053] In some embodiments, machine vision technology can be used to identify misclassified packages and blocked compartments. This involves installing a camera at each compartment, capturing images of packages in real time, and comparing visual features to determine if a package has fallen into the correct compartment or if a blockage has occurred. However, this method requires a separate high-definition camera and image acquisition card for each compartment, which significantly increases hardware procurement and installation costs when there are many compartments (e.g., hundreds). Furthermore, this method is susceptible to factors such as changes in lighting conditions, which may lead to feature extraction failures or misjudgments.

[0054] To address the above problems, this application provides a low-cost solution based on sensor monitoring. The following detailed description, in conjunction with the accompanying drawings, illustrates the package misclassification early warning and anti-blocking control method and system provided in this specification.

[0055] Figure 1 This is an exemplary flowchart of a package misclassification warning and anti-blocking control method according to some embodiments of this specification. In some embodiments, the package misclassification warning and anti-blocking control method can be executed by processing logic, which may include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (instructions running on a processing device to execute hardware simulations), and any combination thereof. In some embodiments, Figure 1 One or more operations in the flowchart of the package misclassification warning and anti-blocking control method shown can be implemented by a processing device and / or a terminal device. For example, the package misclassification warning and anti-blocking control method can be stored in a storage device in the form of a computer program and / or instructions, and invoked and / or executed by the processing device and / or the terminal device.

[0056] Reference Figure 1 The package misclassification warning and anti-blocking control method provided in this application embodiment may include the following steps S111~S115:

[0057] Step S111: When a package enters the sorting loop, the package's classification information, posture information, and current location information are obtained. In some embodiments, step S111 can be performed by the information acquisition module 210 mentioned later.

[0058] In this embodiment, one (or more) correction cameras can be configured at the beginning of the sorting loop (e.g., above the sorting loading table) to acquire package images. It should be noted that the correction camera is a device with image acquisition and correction functions. It can be used to photograph packages entering the beginning of the sorting loop and automatically perform correction processing (e.g., rotation, scaling) on ​​the images to obtain package images that more accurately reflect package classification information, posture information, and other data. In other words, by processing and analyzing these package images, the required package classification information, posture information, and other data can be extracted for subsequent analysis.

[0059] Specifically, in this embodiment, the barcode (e.g., barcode, QR code, etc.), size (e.g., outer contour dimensions), packaging type (e.g., cardboard box, plastic bag, woven bag, etc.), and destination information of the package can be identified based on the package image. Then, the classification information of the package can be determined based on at least one of the barcode, size, packaging type, destination information, and RFID (Radio Frequency Identification) tag information corresponding to the package. The specific implementation of identifying the barcode, size, packaging type, and destination information of the package based on the package image can be considered as prior art and will not be discussed in detail in this specification.

[0060] In some embodiments of this application, the barcode corresponding to the package can be identified, and then relevant data can be queried from the system based on the information fed back by the barcode to obtain information such as its size, weight, packaging type and destination.

[0061] In this embodiment, the classification information of a package can be determined based on at least one of the barcode, size, packaging type, destination information, and the RFID tag information corresponding to the package. For example, in some embodiments, basic data of the package can be retrieved through barcode information, and a preliminary classification can be completed by combining size and packaging type, and then the classification level can be refined using destination information. In some embodiments, a preliminary classification can be performed first using destination information, and then the classification can be refined using size and packaging type. In some embodiments, classification can be performed directly using destination information. In some embodiments, detailed classification data of the package can be directly extracted from the system database using the RFID tag information corresponding to the package, including relevant information such as logistics route, priority level, and special handling requirements, and then classified based on this. In some embodiments, detailed classification data of the package can also be directly extracted from the system database using the barcode information corresponding to the package, such as relevant information such as logistics route, priority level, and special handling requirements, and then classified based on this.

[0062] Furthermore, in this embodiment of the application, the relative position information of the package relative to the sorting loading table can be identified based on the package image, and then the posture information corresponding to the package can be obtained based on the relative position information.

[0063] Specifically, in this embodiment, the sorting loading table refers to the table on which the package is placed before entering the sorting loop. The posture information of the package can be used to reflect the placement state of the package on the sorting loading table (in this application, it mainly refers to the relative position information of the package on the sorting loading table). It should be noted that in this embodiment, the sorting loading table can be cyclically moving (i.e., the sorting loading table can be one of multiple tables included in the sorting loop) or fixed (i.e., the sorting loading table and the sorting loop can be connected by a transmission device, and the package on the sorting loading table can enter the sorting loop through the transmission device).

[0064] It should also be noted that in actual design, to meet the applicability of the sorting loading table, the sorting loading table is usually designed to be relatively large. When packages enter the sorting loop with different placement postures (i.e., different positions on the sorting loading table), there may be different degrees of missorting risk. In the embodiments of this application, by obtaining the posture information of the package, the movement trajectory and possible situations of the package in the subsequent sorting process can be determined more accurately, and the time for the package to reach the target compartment can be estimated more accurately, thereby more accurately determining whether the package has been missorted.

[0065] In the actual sorting process, packages to be sorted can be placed on the sorting platform manually or automatically (e.g., conveyor belts, robotic arms, etc.). Then, one or more correction cameras positioned above the sorting platform capture images of the packages. Further, image recognition technology can be used to capture the specific position of the package on the platform, thereby obtaining the package's relative position information with respect to the sorting platform. Based on this relative position information, the package's posture information can then be derived. Specifically, in some embodiments of this application, the posture information of the package can be represented in the form of relative position information.

[0066] Furthermore, in this embodiment, the current location information of the package can be determined based on the distribution location information (e.g., distribution location coordinates) of the sorting loading platform in the sorting loop. In some embodiments of this application, a two-dimensional or three-dimensional coordinate system can be established for the sorting loop, and then the location coordinates of the sorting loading platform in the coordinate system can be determined, and the current location information of the package can be obtained based on the location coordinates corresponding to the sorting loading platform.

[0067] Step S112: Determine the target compartment corresponding to the package based on the classification information, and determine the target location information based on the location of the target compartment. In some embodiments, step S112 can be performed by the target compartment determination module 220 mentioned below.

[0068] In this embodiment of the application, the target compartment corresponding to the package can be determined based on the classification information obtained from the aforementioned process, and the target location information can be determined based on the location of the target compartment.

[0069] Specifically, in the embodiments of this application, packages of different categories can be transferred to different compartments during the sorting process, and these different compartments are located at different positions in the sorting loop. Based on this, in some embodiments of this application, the correspondence between different classification information and compartments can be pre-configured, and then the target compartment to which the package will be transported can be determined according to the correspondence and the classification information determined for the package in the aforementioned process.

[0070] Meanwhile, in the embodiments of this application, the target location information can be determined based on the location (absolute location) of the target compartment. As an example only, in some embodiments of this application, a two-dimensional or three-dimensional coordinate system can be established for the sorting loop, and then the location of each compartment can be represented by coordinates.

[0071] Step S113: Predict the time range within which the package triggers the landing sensor of the target compartment based on the attitude information, current position information, and target position information. In some embodiments, step S113 may be performed by the time prediction module 230 mentioned below.

[0072] In this embodiment, a drop sensor can be configured in each compartment to detect whether a package has fallen into the compartment. The drop sensor can be a photoelectric sensor (such as a through-beam photoelectric sensor or a reflective photoelectric sensor, which determines whether a package has fallen in by detecting changes in the blocking or reflection of light). When a package passes through the compartment and blocks or reflects light, the drop sensor will immediately generate a corresponding electrical signal and transmit the signal to the processing device for further processing and analysis.

[0073] In this embodiment, the time range within which the package triggers the drop sensor of the target compartment can be predicted based on the attitude information, current position information, and target position information obtained from the aforementioned process. For example, in some embodiments, a standard reference time range required for the package to reach the target compartment can be predicted based on the current position information, the target position information, and the running speed information of the sorting loop; then, combined with the attitude information, the standard reference time range for the package to reach the target compartment is corrected to obtain the time range within which the package triggers the drop sensor of the target compartment.

[0074] In this embodiment, the standard reference time range can be understood as the normal time range required for a package to be transferred from the sorting platform to the target compartment without any unexpected events during the sorting process. This standard reference time range can be obtained by extending a calculated specific time value by a preset value (e.g., 1 second). However, since the sorting platform is usually designed to be large, the time it takes for a package to be transferred from the sorting platform to the sorting loop, and the time it takes for the package to fall into the compartment, are both affected to some extent by the aforementioned posture information. Therefore, when a package enters the sorting loop with different placement postures (i.e., at different positions on the sorting platform), there may be different degrees of missorting risk and placement time error. To address this problem, this embodiment uses the posture information obtained in the above process to correct the standard reference time range for the package to reach the target compartment, thereby obtaining the time range within which the package triggers the placement sensor of the target compartment.

[0075] For example, in some embodiments of this application, the midpoint of the standard reference time range can be used as the center, and the correction magnitude can be determined based on the error time calculated based on the attitude information, thereby determining a time interval as the time range for the package to trigger the landing sensor of the target grid.

[0076] In some embodiments, considering that the error time may be related to multiple factors such as the aforementioned posture information, size, weight, and packaging type, in some embodiments of this application, to ensure the accuracy and reliability of the calculated error time, the above factors can be calculated together. For example, an error time calculation model can be established, taking multiple influencing factors such as posture information, size, weight, and packaging type as input parameters, and outputting the corresponding error time. Furthermore, the standard reference time range can be corrected based on the error time to obtain a more accurate time range for the package to trigger the drop sensor of the target compartment.

[0077] In some embodiments of this application, the midpoint of the standard reference time range can be used as the center, and the correction magnitude can be determined based on the error time to establish a new time interval as the time range for the package to trigger the landing sensor of the target compartment. Specifically, in some embodiments of this application, a correction time range can be determined based on the error time and the midpoint of the standard reference time range, and then the union of the standard reference time range and the correction time range can be taken as the time range for the package to trigger the landing sensor of the target compartment.

[0078] For example, assuming the standard reference time range is T1 to T2, with a midpoint time of Tm = (T1 + T2) / 2, and the error time obtained through the error time calculation model is ΔT, then the corrected time range can be expressed as (T1 + T2) / 2 to (T1 + T2) / 2 + ΔT. If (T1 + T2) / 2 + ΔT is greater than T2, then the time range in which the package triggers the cell sensor at the target cell can be determined as T1 to (T1 + T2) / 2 + ΔT. Conversely, if (T1 + T2) / 2 + ΔT is less than or equal to T2, then the time range in which the package triggers the cell sensor at the target cell can be determined as T1 to T2. In some embodiments, the aforementioned error time ΔT can be negative, and the corresponding time range determination method can refer to the foregoing content, which will not be repeated here.

[0079] It should be noted that the above error time calculation model can be trained in the following way:

[0080] First, a large amount of historical data related to package posture, size, weight, and packaging type is collected, along with the corresponding actual error time (obtained by calculating the difference between the midpoint of the standard reference time range and the actual package placement time). Then, this historical data is used as training samples, with various package features (i.e., the aforementioned posture, size, weight, and packaging type) as input features and their corresponding actual error times as the target output. Further, a suitable machine learning algorithm or deep learning model architecture, such as a neural network model, can be selected, and the model can be trained using the aforementioned training samples.

[0081] Specifically, during training, the model's parameters can be continuously adjusted, gradually reducing the difference between the predicted error time and the actual error time. It can be understood that through multiple iterations of training with a large number of training samples, the model can achieve better performance metrics, and the resulting model can then be used as an error time calculation model for calculating the aforementioned error time.

[0082] It should also be noted that in some embodiments of this application, the operating speed of the sorting loop may be non-uniform (e.g., stopped or decelerated midway due to manual intervention), and the aforementioned standard reference time range can be corrected in real time according to the actual speed of the sorting loop. The time range at which the package triggers the drop sensor of the target compartment can be calculated based on the corrected standard reference time range and the error time obtained by processing the aforementioned error time calculation model.

[0083] Step S114: After sending a sorting command for the package to the sorting equipment, the sorting sensor signals of the target compartment and its adjacent compartments are monitored in real time. In some embodiments, step S114 may be performed by the signal monitoring module 240 mentioned below.

[0084] In this embodiment of the application, a drop command can be sent from the sorting equipment to the sorting table (one of the tables in the sorting loop) where the package is located, so as to control the sorting table to transport the package in the direction of the target slot (for example, by driving a conveyor belt), so that the package falls into the corresponding slot.

[0085] In this embodiment of the application, after sending a sorting command for the package to the sorting equipment, the sorting sensor signals of the target sorting slot and its adjacent slots can be monitored in real time. The adjacent slots can refer to the N slots adjacent to the target sorting slot (including the N slots before and after; N=1, 2, 3...).

[0086] In some embodiments of this application, the drop sensor signal corresponding to each grid can be transmitted in the form of a specific level signal or data packet. When a package falls into the grid, the corresponding drop sensor will detect the package and generate a corresponding signal change, and transmit the signal to the signal monitoring module.

[0087] It is understood that, in this embodiment, real-time monitoring of the parcel sorting sensor signals allows for timely understanding of the parcel sorting status, providing accurate data for subsequent missorting warnings and anti-blocking control. Furthermore, it should be noted that, in this embodiment, by sending a parcel sorting command to the sorting equipment and then monitoring the parcel sorting sensor signals of the target compartment and its adjacent compartments in real time, resource waste and data redundancy caused by continuous and full-compartment data collection can be avoided. This improves the system's operational efficiency and data processing capabilities to a certain extent, making the entire parcel missorting warning and anti-blocking control process more efficient and reliable.

[0088] Step S115: If it is detected that the target compartment does not trigger its parcel placement sensor signal within the predicted time range, and the adjacent compartment has no predetermined parcels to be placed within or near the time range, but triggers its parcel placement sensor signal within or near the time range, then it is determined that the parcel has been misplaced, and a misplacement warning is triggered. In some embodiments, step S115 can be executed by the misplacement determination module 250 mentioned below.

[0089] In this embodiment, if the target compartment fails to trigger its placement sensor signal within the predicted time range, it indicates that the package has not fallen into the target compartment it should have fallen into. Since the time range for triggering the target compartment's placement sensor, calculated through the above process, has considered almost all factors affecting the package's arrival at the target compartment and includes a certain margin of error (i.e., the aforementioned pre- and post-extension preset values), when the target compartment fails to trigger its placement sensor signal within the predicted time range, it indicates that the package has most likely been misclassified to another compartment. In this case, for further verification, it is possible to detect if the adjacent compartments corresponding to the target compartment trigger their placement sensor signals within or near the predicted time range.

[0090] Specifically, in this embodiment, if it is detected that the target compartment does not trigger its parcel placement sensor signal within the predicted time range, and simultaneously, the adjacent compartment corresponding to the target compartment triggers its parcel placement sensor signal within or near the time range (specifically determined by the number of N adjacent compartments before and after the target compartment; the larger the value of N, the larger the specific time range referred to by the "near the time range"), and the compartment that triggered the parcel placement sensor signal did not originally have a pre-determined parcel placed within or near the time range, then it can be determined that the parcel has been misplaced, and a misplacement warning can be triggered. As an example only, in this embodiment, the misplacement warning can be presented in various forms such as an audible alarm, flashing lights, or sending notification information to the operator's terminal.

[0091] In some embodiments of this application, in order to further ensure the accuracy of the above-mentioned missorting judgment results, when it is detected that there are predetermined parcels to be sorted in adjacent compartments within adjacent time periods, the sorting time of the later parcel to be sorted can be delayed (for example, the sorting instruction can be sent to its corresponding sorting table after the sorting loop has moved one or more times).

[0092] It should be noted that, in this embodiment of the application, by delaying the arrival time of the later package when it is detected that there are predetermined packages to be placed in adjacent compartments within adjacent time periods, the interference that may occur when there are predetermined packages to be placed in adjacent compartments within adjacent time periods can be reduced to a certain extent.

[0093] In some embodiments of this application, the number of packages falling into a compartment can also be detected by a drop sensor configured in each compartment. In some embodiments, if the target compartment does not trigger its drop sensor signal within the predicted time range, the triggering status of the drop sensor signals of adjacent compartments and the number of packages dropped into them can be further determined. Specifically, if the target compartment does not trigger its drop sensor signal within the predicted time range, and the adjacent compartment triggers its drop sensor signal within or near the time range, and the number of packages falling into the adjacent compartment is greater than the predetermined number of packages dropped into within or near the time range, it can also be determined that the packages have been misclassified, and a misclassification warning can be triggered in this case.

[0094] In some embodiments of this application, to facilitate the sorting and locating of missorted packages, a buffer area connected to each sorting slot can be set up. This buffer area can be used to temporarily store the latest package that has been placed in a slot. When a missorted package is detected flowing into any slot, the latest package stored in the corresponding buffer area can be re-flowed into the sorting loop, thereby automatically achieving secondary sorting.

[0095] In some embodiments of this application, the signal status of each grid cell sensor can be monitored in real time. Then, when the sensor signal of a certain grid cell is detected to be continuously triggered and exceeds a preset first time threshold (e.g., 5 seconds), it is determined that a grid cell blockage has occurred, and a blockage alarm is triggered. In some embodiments, this process can be performed by the blockage determination module 260 mentioned below.

[0096] In this embodiment, the grid blocking alarm can take the same or different form as the aforementioned misclassification warning. Specifically, in this embodiment, the grid blocking alarm can take any one of the following forms: sound alarm, flashing light, or sending notification information to the operator's terminal.

[0097] In this embodiment of the application, to avoid more serious consequences after package misclassification or blockage occurs, the placement time of related packages awaiting placement can be delayed when any misclassified package is detected flowing into a compartment or a blockage occurs. The related packages awaiting placement refer to packages that are scheduled to be placed into the compartment where the misclassification or blockage has occurred.

[0098] As an example only, in some embodiments of this application, when a missorted package is detected flowing into a certain compartment or a compartment blockage occurs, the relevant package to be sorted can be controlled to wait for a preset number of loops on the sorting loop. If the compartment is still blocked after the maximum number of loops, the package can be unloaded to a re-sorting compartment (the re-sorting compartment is one or more compartments in the sorting loop).

[0099] Figure 2 This is an exemplary flowchart illustrating a package misclassification warning and anti-blocking control method according to other embodiments of this specification. (Refer to...) Figure 2 In some embodiments of this application, the package misclassification warning and anti-blocking control method may further include the following steps S121~S124:

[0100] Step S121: Obtain historical sorting data and construct a training sample set, wherein the historical sorting data includes the classification information, posture information and corresponding sorting result information of historical packages.

[0101] In this embodiment, historical sorting data can be obtained by collecting operational data of the sorting system over a past period. The classification and posture information corresponding to the historical package can be obtained using the methods described above, and the sorting result information corresponding to the historical package can be obtained using the mis-sorting determination process described above. The specific methods for obtaining these information will not be discussed in detail here.

[0102] Step S122: Train a misclassification risk prediction model based on the training sample set. The misclassification risk prediction model is configured to output the corresponding misclassification risk assessment result based on the classification information and posture information of the packages entering the sorting loop.

[0103] Step S123: For packages entering the sorting loop, the trained missorting risk prediction model is used to determine the corresponding missorting risk assessment result.

[0104] In this embodiment, after training the misclassification risk prediction model using the aforementioned training sample set, it gains the ability to predict misclassification risk based on the package's classification and orientation information. Specifically, when a package enters the sorting loop, its corresponding classification and orientation information can be obtained and input into the trained misclassification risk prediction model. Furthermore, the misclassification risk prediction model can perform calculations and analyses based on the input information, outputting a misclassification risk assessment result for the package.

[0105] Step S124: Based on the misclassification risk assessment results, implement a predictive processing strategy for high-risk packages.

[0106] In this embodiment, the misclassification risk assessment result obtained through the above process reflects the probability of misclassification of the corresponding package. In some embodiments, packages can be divided into different levels based on the probability of misclassification. For example, packages can be divided into three levels: high risk, medium risk, and low risk. Among them, for high-risk packages, that is, packages whose misclassification risk assessment results show a high probability of misclassification, a predictive processing strategy can be implemented.

[0107] In the embodiments of this application, the predictive processing strategy may include, but is not limited to, any one of the following: posture correction (e.g., adjusting the placement posture of the package on the sorting table by means of a robotic arm), re-sorting (e.g., separating high-risk packages from the normal sorting loop for secondary sorting, manual verification, or using more precise sorting equipment for secondary sorting), and manual intervention (e.g., issuing prompts to remind relevant personnel to correct the posture of packages entering the sorting loop).

[0108] Figure 3 This is an exemplary training flowchart of a misclassification risk prediction model according to some embodiments of this specification. The following is in conjunction with... Figure 3 The training process of the misclassification risk prediction model involved in the embodiments of this application is briefly described. (Refer to...) Figure 3 In some embodiments of this application, the training process of the above-mentioned misclassification risk prediction model may include the following steps S101~S103:

[0109] Step S101: Using the classification information and posture information corresponding to historical packages as input features, and the misclassification results reflected in the sorting results information corresponding to historical packages as labels, a training sample set is constructed.

[0110] Step S102: The initial machine learning model is trained using the training sample set, and the parameters of the initial machine learning model are continuously adjusted through an optimization algorithm so that the error between the misclassification risk assessment result output by the model and the label gradually decreases.

[0111] In this embodiment of the application, the initial machine learning model may include any one of a neural network model, a support vector machine model, or a random forest model, and the optimization algorithm includes the gradient descent algorithm.

[0112] Step S103: When the error between the misclassification risk assessment result output by the model and the label is less than a preset threshold, or when the training reaches the preset maximum number of iterations, training is stopped, and the trained misclassification risk prediction model is obtained.

[0113] Specifically, in this embodiment of the application, a loss function can be constructed based on the error between the misclassification risk assessment result output by the model and the label (for example, the corresponding loss function can be obtained by calculating the distance between the output misclassification risk assessment result and the label), and then the model parameters can be optimized by minimizing the loss function.

[0114] When the loss function value is less than the preset threshold (which can be set according to actual needs), it indicates that the error between the model's output misclassification risk assessment result and the label is small enough, and the model performance has reached a good level. At this point, training can be terminated, and a well-trained misclassification risk prediction model can be obtained.

[0115] In some embodiments of this application, when the training reaches the preset maximum number of iterations (at which point continuing training may not bring significant performance improvement, but will instead increase computational cost and time), training can also be stopped to obtain a trained misclassification risk prediction model.

[0116] Further details regarding the construction and training (i.e., model optimization process) of this misclassification risk prediction model can be considered as existing technology and will not be elaborated upon in this specification.

[0117] It should be noted that, in this embodiment of the application, by using the above-mentioned misclassification risk prediction model to assess the misclassification risk of packages entering the sorting loop, and then executing a predictive processing strategy based on the obtained misclassification risk assessment results, the probability of misclassification can be effectively reduced, thereby improving the sorting efficiency of packages to a certain extent.

[0118] Figure 4This is a schematic diagram of a package missorting warning and anti-blocking control system according to some embodiments of this specification. In some embodiments, Figure 4 The package missorting warning and anti-blocking control system 200 shown can be implemented in software and / or hardware. For example, it can be configured in the form of software and / or hardware to process equipment and / or terminal equipment to automatically identify missorting and blockage phenomena in the package sorting process and issue corresponding warnings.

[0119] Reference Figure 4 The package misclassification warning and anti-blocking control system 200 provided in this application embodiment may include an information acquisition module 210, a target compartment determination module 220, a time prediction module 230, a signal monitoring module 240, and a misclassification judgment module 250.

[0120] The information acquisition module 210 can be used to acquire the package's classification information, posture information, and current location information when the package enters the sorting loop.

[0121] The target compartment determination module 220 can be used to determine the target compartment corresponding to the package based on the classification information, and to determine the target location information based on the location of the target compartment.

[0122] The time prediction module 230 can be used to predict the time range within which the package triggers the landing sensor of the target compartment based on the attitude information, current position information, and target position information.

[0123] The signal monitoring module 240 can be used to monitor the drop-off sensor signals of the target compartment and its adjacent compartments in real time after sending a drop-off command to the sorting equipment for the package.

[0124] The misclassification determination module 250 can be used to determine that the package has been misclassified when it is detected that the target compartment has not triggered its parcel sensor signal within the predicted time range, and the adjacent compartment has no predetermined parcels within or near the time range, but has triggered its parcel sensor signal within or near the time range. At the same time, it triggers a misclassification warning.

[0125] Continue to refer to Figure 4 In some embodiments, the package missorting warning and anti-blocking control system 200 may further include a blockage determination module 260, which can be used to: monitor the signal status of the sensor in each compartment in real time; when the sensor signal of a certain compartment is continuously in the triggered state and exceeds a preset first time threshold, determine that the compartment is blocked and trigger a blockage alarm.

[0126] For more details about the above modules, please refer to other parts of this manual (e.g., Figures 1-3(Parts and related descriptions), which will not be repeated here.

[0127] It should be understood that Figure 4 The package misclassification warning and anti-blocking control system 200 and its modules shown can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated hardware. Those skilled in the art will understand that the above-described methods and systems can be implemented using computer-executable instructions and / or included in processor control code, for example, on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The systems and modules of this specification can be implemented not only by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., but also by software, for example, executed by various types of processors, or by a combination of the above-described hardware circuits and software (e.g., firmware).

[0128] It should be noted that the above description of the package misclassification warning and anti-blocking control system 200 is provided for illustrative purposes only and is not intended to limit the scope of this specification. It will be understood that those skilled in the art can, based on the description in this specification, arbitrarily combine the various modules, or construct subsystems connected to other modules, without departing from this principle. For example, Figure 4 The information acquisition module 210, target grid determination module 220, time prediction module 230, signal monitoring module 240, misclassification judgment module 250, and grid blocking judgment module 260 described herein can be different modules within a single system, or a single module can implement the functions of two or more of the aforementioned modules. Such variations are all within the scope of protection of this specification. In some embodiments, the aforementioned modules can be part of a processing device and / or a terminal device.

[0129] In summary, the beneficial effects that the embodiments of this specification may bring include, but are not limited to:

[0130] (1) In the package missorting warning and anti-blocking control method and system provided in some embodiments of this specification, the time range of the package triggering the landing sensor of the target compartment is predicted according to the package classification information, posture information and current location information. Then, after sending the landing command for the package to the sorting equipment, the landing sensor signal of the target compartment and its adjacent compartments is monitored in real time. Finally, the triggering status of the landing sensor signal of the target compartment and its adjacent compartments within the time range is used to determine whether the package is missorted. The monitoring of package missorting can be realized at low cost, which greatly reduces the implementation cost of package missorting and blockage monitoring.

[0131] (2) In the package missorting warning and anti-blocking control method and system provided in some embodiments of this specification, by sending a sorting command for the package to the sorting equipment and monitoring the sorting sensor signals of the target compartment and its adjacent compartments in real time, the waste of resources and data redundancy caused by full-time collection and full-compartment data collection can be avoided, thereby improving the system's operating efficiency and data processing capabilities to a certain extent, making the entire package missorting warning and anti-blocking control process more efficient and reliable.

[0132] (3) In the package missorting warning and anti-blocking control method and system provided in some embodiments of this specification, by setting up a buffer area connected to each grid to temporarily store the latest package that has been placed in the grid, it is not only convenient to sort and find packages that have been missorted, but also to automatically realize secondary sorting when a missorted package is detected flowing into any grid.

[0133] (4) In the package misclassification warning and anti-blocking control method and system provided in some embodiments of this specification, by delaying the placement time of the relevant package to be placed in the grid when a misclassified package is detected to flow into any grid or a grid blockage occurs, it is possible to avoid further serious consequences after the package misclassification or grid blockage occurs.

[0134] (5) In the package missorting early warning and anti-blocking control method and system provided in some embodiments of this specification, the package entering the sorting loop is assessed for missorting risk by using a missorting risk prediction model, and then a predictive processing strategy is executed based on the obtained missorting risk assessment results. This can effectively reduce the probability of missorting and thus improve the sorting efficiency of packages to a certain extent.

[0135] It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects may be any one or a combination of the above, or any other possible beneficial effects.

[0136] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification and therefore remain within the spirit and scope of the exemplary embodiments described herein.

[0137] Furthermore, this specification uses specific terms to describe embodiments thereof. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Moreover, certain features, structures, or characteristics in one or more embodiments of this specification can be appropriately combined.

[0138] Furthermore, those skilled in the art will understand that various aspects of this specification can be described and illustrated in several patentable ways or situations, including any new and useful combination of processes, machines, products, or substances, or any new and useful improvements thereof. Accordingly, various aspects of this specification can be implemented entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. All of the above hardware or software may be referred to as a “data block,” “module,” “engine,” “unit,” “component,” or “system.” Furthermore, various aspects of this specification may be represented as a computer product located on one or more computer-readable media, including computer-readable program code.

[0139] Computer storage media may contain a propagated data signal containing computer program code, for example, on baseband or as part of a carrier wave. This propagated signal may take various forms, including electromagnetic, optical, and suitable combinations thereof. Computer storage media can be any computer-readable medium other than a computer-readable storage medium, which can be connected to an instruction execution system, apparatus, or device to enable communication, propagation, or transmission of a program for use. The program code located on the computer storage medium can be propagated through any suitable medium, including radio, cable, fiber optic cable, RF, or similar media, or any combination of the above media.

[0140] The computer program code required for the operation of each part of this manual can be written in any one or more programming languages, including object-oriented programming languages ​​such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, etc.; conventional procedural programming languages ​​such as C, Visual Basic, Fortran2003, Perl, COBOL2002, PHP, ABAP; dynamic programming languages ​​such as Python, Ruby, and Groovy; or other programming languages. This program code can run entirely on the user's computer, or as a standalone software package on the user's computer, or partially on the user's computer and partially on a remote computer, or entirely on a remote computer or processing device. In the latter case, the remote computer can be connected to the user's computer through any network, such as a local area network (LAN) or wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or used as a service such as Software as a Service (SaaS).

[0141] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this specification are not intended to limit the order of the processes and methods described herein. Although various examples have been discussed in the foregoing disclosure of some embodiments of the invention that are currently considered useful, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the spirit and scope of the embodiments described herein. For example, while the system components described above can be implemented by hardware devices, they can also be implemented solely by software solutions, such as installing the described system on existing processing devices or mobile devices.

[0142] Similarly, it should be noted that, in order to simplify the description disclosed herein and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of embodiments in this specification may sometimes combine multiple features into a single embodiment, drawing, or description thereof. However, this method of disclosure does not imply that the subject matter of this specification requires more features than those mentioned in the claims. In fact, the embodiments contain fewer features than all the features of a single embodiment disclosed above.

[0143] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A method for early warning and prevention of package misclassification, characterized in that, include: When a package enters the sorting loop, its classification information, posture information, and current location information are obtained. The target compartment corresponding to the package is determined based on the classification information, and the target location information is determined based on the location of the target compartment. Based on the attitude information, current position information, and target position information, predict the time range within which the package triggers the drop sensor of the target compartment. After sending a sorting command for the package to the sorting equipment, the sorting sensor signals of the target compartment and its adjacent compartments are monitored in real time. If it is detected that the target compartment does not trigger the corresponding parcel drop sensor signal within the predicted time range, and there is no predetermined parcel drop in the adjacent compartment within or near the time range, but the parcel drop sensor signal corresponding to the adjacent compartment is triggered within or near the time range, then it is determined that the parcel has been missorted, and a missorting warning is triggered. The step of predicting the time range for the package to trigger the landing sensor of the target compartment based on the attitude information, current position information, and target position information includes: Based on the current location information, the target location information, and the operating speed information of the sorting loop, predict the standard reference time range required for the package to reach the target compartment; Based on the attitude information, the standard reference time range for the package to reach the target compartment is corrected to obtain the time range for the package to trigger the compartment sensor of the target compartment; wherein, the correction of the standard reference time range for the package to reach the target compartment includes: taking the midpoint of the standard reference time range as the center, and then determining the correction magnitude based on the error time calculated based on the attitude information, thereby determining a time interval as the time range for the package to trigger the compartment sensor of the target compartment.

2. The parcel misclassification early warning and anti-blocking control method as described in claim 1, characterized in that, The acquisition of the package's classification information, posture information, and current location information includes: Image of the package is acquired using a correction camera; Based on the package image, the barcode, size, packaging type, and destination information of the package are identified. Then, based on at least one of the barcode, size, packaging type, destination information, and the RFID tag information of the package, the classification information of the package is determined. Based on the package image, the relative position information of the package relative to the sorting loading table is identified, and then the posture information corresponding to the package is obtained based on the relative position information; Based on the distribution location information of the sorting loading platform in the sorting loop, the current location information of the package is determined.

3. The parcel misclassification early warning and anti-blocking control method as described in claim 1, characterized in that, Each compartment is equipped with a buffer area connected to it, which is used to temporarily store the latest parcel that has been placed in the compartment. The method also includes: when a missorted parcel is detected flowing into any compartment, the latest parcel that has been placed in the corresponding buffer area is made to flow back into the sorting loop.

4. The parcel misclassification early warning and anti-blocking control method as described in claim 1, characterized in that, The method further includes: Real-time monitoring of the signal status of the sensor at each grid opening; When the sensor signal of a certain grid is continuously triggered and exceeds the preset first time threshold, it is determined that the grid is blocked and a grid blockage alarm is triggered.

5. The package misclassification early warning and anti-blocking control method as described in claim 4, characterized in that, The method further includes: when a misclassified package is detected flowing into any compartment or a blockage occurs, delaying the placement time of the relevant package to be placed in the compartment.

6. The parcel misclassification early warning and anti-blocking control method as described in any one of claims 1 to 5, characterized in that, The method further includes: Acquire historical sorting data and construct a training sample set, wherein the historical sorting data includes the classification information, posture information and corresponding sorting result information of historical packages; A misclassification risk prediction model is trained based on the training sample set. The misclassification risk prediction model is configured to output the corresponding misclassification risk assessment result based on the classification information and posture information of the packages entering the sorting loop. For packages entering the sorting loop, the trained misclassification risk prediction model is used to determine the corresponding misclassification risk assessment result. Based on the missorting risk assessment results, a predictive handling strategy is implemented for high-risk packages; wherein, the predictive handling strategy includes any one of posture correction, manual intervention, and re-sorting.

7. The package misclassification early warning and anti-blocking control method as described in claim 6, characterized in that, The misclassification risk prediction model is trained in the following manner: Using the classification information and posture information corresponding to the historical packages as input features, and the misclassification phenomenon recognition results reflected by the sorting result information corresponding to the historical packages as labels, a training sample set is constructed; The initial machine learning model is trained using the training sample set, and the parameters of the initial machine learning model are continuously adjusted through an optimization algorithm so that the error between the misclassification risk assessment result output by the model and the label gradually decreases; wherein, the initial machine learning model includes any one of a neural network model, a support vector machine model, or a random forest model, and the optimization algorithm includes the gradient descent algorithm; When the error between the model's output misclassification risk assessment result and the label is less than a preset threshold, or when the training reaches the preset maximum number of iterations, training stops, and the trained misclassification risk prediction model is obtained.

8. A package missorting early warning and anti-blocking control system, characterized in that, include: The information acquisition module is used to acquire the package's classification information, posture information, and current location information when the package enters the sorting loop. The target compartment determination module is used to determine the target compartment corresponding to the package based on the classification information, and to determine the target location information based on the location of the target compartment; The time prediction module is used to predict the time range within which the package triggers the cell sensor of the target compartment based on the attitude information, current position information, and target position information. The signal monitoring module is used to monitor the drop-off sensor signals of the target compartment and its adjacent compartments in real time after sending a drop-off command to the sorting equipment for the package. The misclassification determination module is used to determine that the package has been misclassified when it is detected that the target compartment has not triggered the corresponding parcel sensor signal within the predicted time range, and there is no predetermined parcel in the adjacent compartment within or near the time range, but the parcel sensor signal corresponding to the adjacent compartment has been triggered within or near the time range; at the same time, it triggers a misclassification warning. Specifically, the time prediction module is used for: Based on the current location information, the target location information, and the operating speed information of the sorting loop, predict the standard reference time range required for the package to reach the target compartment; Based on the attitude information, the standard reference time range for the package to reach the target compartment is corrected to obtain the time range for the package to trigger the compartment sensor of the target compartment; wherein, the correction of the standard reference time range for the package to reach the target compartment includes: taking the midpoint of the standard reference time range as the center, and then determining the correction magnitude based on the error time calculated based on the attitude information, thereby determining a time interval as the time range for the package to trigger the compartment sensor of the target compartment.

9. The package missorting early warning and anti-blocking control system as described in claim 8, characterized in that, The system also includes a grid blocking determination module, which is used for: Real-time monitoring of the signal status of the sensor at each grid opening; When the sensor signal of a certain grid is continuously triggered and exceeds the preset first time threshold, it is determined that the grid is blocked and a grid blockage alarm is triggered.

Images