Sorting control system and method for sorting machine

Abstract

The embodiments of the present specification provide a sorting control system and method for a sorting machine, the method comprising obtaining cargo characteristics of target cargo and conveying characteristics of a conveyor belt, the target cargo being cargo that is currently located on the conveyor belt and whose distance from the swing belt is less than a preset distance threshold, the cargo characteristics comprising the height, weight and placement posture of the target cargo, and the conveying characteristics comprising the conveying speed of the conveyor belt and cargo interval; based on the cargo characteristics and conveying characteristics, determining an estimated initial velocity and an estimated attitude angle of the target cargo when it arrives at the swing belt; based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle of the target cargo and the target sorting direction of the target cargo, determining a roller parameter matrix of the swing belt.

Description

Technical Field

[0001] The present invention relates to the technical field of item sorting, and in particular to a sorting control system and method for a sorting machine. Background Art

[0002] Logistics distribution centers usually use sorting machines to sort and transport goods continuously and in large quantities. When the sorting machine is sorting and transporting, the sorting machine (for example, a swing wheel sorting machine) needs to adjust the rotation direction of the roller to control the conveying direction of the goods above the roller. Due to the different sizes and weights of different goods, when the roller runs according to the preset parameters, some goods may overturn or roll over, or some goods may be delayed on the conveyor belt, resulting in the accumulation of multiple goods, which ultimately causes some goods to fail to be transported to the designated location.

[0003] Therefore, it is necessary to provide a sorting control system and method for a sorting machine, which can solve the above-mentioned problems that may occur and ensure that the goods are sorted and transported to the designated location. Summary of the invention

[0004] One or more embodiments of the present specification provide a sorting control method for a sorting machine, the method comprising: obtaining cargo characteristics of a target cargo and conveying characteristics of a conveyor belt, the target cargo being cargo that is currently located on the conveyor belt and whose distance from the swing belt is less than a preset distance threshold, the cargo characteristics comprising the height, weight and placement posture of the target cargo, and the conveying characteristics comprising the conveying speed and cargo interval of the conveyor belt; determining an estimated initial velocity and an estimated attitude angle of the target cargo when it arrives at the swing belt based on the cargo characteristics and the conveying characteristics; determining a roller parameter matrix of the swing belt based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle and the target sorting direction of the target cargo.

[0005] One or more embodiments of the embodiments of the present specification provide a sorting control system for a sorting machine, the system comprising an acquisition module, a first determination module and a second determination module; the acquisition module is configured to acquire cargo characteristics of a target cargo and conveying characteristics of a conveyor belt, the target cargo is cargo that is currently located on the conveyor belt and whose distance from the swing belt is less than a preset distance threshold, the cargo characteristics include the height, weight and placement posture of the target cargo, and the conveying characteristics include the conveying speed and cargo interval of the conveyor belt; the first determination module is configured to determine an estimated initial velocity and an estimated attitude angle of the target cargo when it arrives at the swing belt based on the cargo characteristics and the conveying characteristics; the second determination module is configured to determine a roller parameter matrix of the swing belt based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle and the target sorting direction of the target cargo.

[0006] One or more embodiments of the present specification provide a sorting control device for a sorting machine, the device comprising at least one processor and at least one memory, the at least one memory being used to store computer instructions; the at least one processor being used to execute a sorting control method for a sorting machine as described in any embodiment of the present specification.

[0007] One or more embodiments of the present specification provide a computer-readable storage medium, wherein the storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes a sorting control method for a sorting machine as described in any embodiment of the present specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:

[0009] Figure 1 is a module diagram of a sorting control system of a sorting machine according to some embodiments of this specification;

[0010] Figure 2 is an exemplary flow chart of a sorting control method for a sorting machine according to some embodiments of this specification;

[0011] Figure 3 is an exemplary schematic diagram of a sorting control method for a sorting machine according to some embodiments of this specification;

[0012] Figure 4 is an exemplary schematic diagram of a stability prediction model according to some embodiments of the present specification. DETAILED DESCRIPTION

[0013] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of this specification. For ordinary technicians in this field, this specification can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0014] It should be understood that the "system", "device", "unit" and / or "module" used herein are a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0015] As shown in this specification and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "comprise" and "include" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0016] Flowcharts are used in this specification to illustrate the operations performed by the system according to the embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed precisely in order. Instead, the steps may be processed in reverse order or simultaneously. At the same time, other operations may also be added to these processes, or one or more operations may be removed from these processes.

[0017] In order to avoid various problems that may occur during the sorting and transportation of goods, achieve accurate and efficient sorting of goods, and ensure that the goods are sorted and transported to designated locations, some embodiments of the present specification provide a sorting control system and method for a sorting machine, the method comprising obtaining cargo characteristics of target goods and conveying characteristics of a conveyor belt, the target goods being goods that are currently on the conveyor belt and whose distance from the balance belt is less than a preset distance threshold; determining an estimated initial velocity and an estimated attitude angle of the target goods when they arrive at the balance belt based on the cargo characteristics and the conveying characteristics; determining a roller parameter matrix of the balance belt based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle of the target goods and the target sorting direction of the target goods.

[0018] Figure 1 It is a module diagram of a sorting control system of a sorting machine according to some embodiments of this specification.

[0019] like Figure 1 As shown, in some embodiments, the sorting machine sorting control system 100 may include an acquisition module 110 , a first determination module 120 and a second determination module 130 .

[0020] In some embodiments, the acquisition module 110 is configured to acquire cargo characteristics of the target cargo and conveying characteristics of the conveyor belt.

[0021] In some embodiments, the first determination module 120 is configured to determine the estimated initial velocity and the estimated attitude angle of the target cargo when it arrives at the balance belt based on the cargo characteristics and the transportation characteristics.

[0022] In some embodiments, the second determination module 130 is configured to determine the roller parameter matrix of the balance belt based on cargo characteristics, estimated initial velocity, estimated attitude angle, and target sorting direction of the target cargo.

[0023] In some embodiments, the second determination module 130 is further configured to determine the sorting abnormality probability of the target cargo based on cargo characteristics, estimated initial velocity, estimated attitude angle and target sorting direction; in response to the sorting abnormality probability being less than a preset threshold, determine the roller parameter matrix based on cargo characteristics, estimated initial velocity, estimated attitude angle and target sorting direction.

[0024] In some embodiments, in response to the sorting abnormality probability being greater than a preset threshold, the second determination module 130 is further configured to: generate multiple candidate roller parameter matrices according to the target sorting direction; for each candidate roller parameter matrix, based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle and the candidate roller parameter matrix, determine the stability of the target cargo under the candidate roller parameter matrix through a stability prediction model, where the stability prediction model is a machine learning model; determine the roller parameter matrix based on the stabilities corresponding to each of the multiple candidate roller parameter matrices.

[0025] In some embodiments, the sorting machine sorting control system 100 may further include a feedback module ( Figure 1 In some embodiments, the feedback module is configured to determine the instability distribution by calculating the instability values ​​of a plurality of sorted goods within a preset historical period; in response to the instability distribution meeting a preset condition, generate and send a fault warning, and / or adjust the conveying speed of the conveyor belt.

[0026] In some embodiments, for each sorted cargo, the feedback module is further configured to determine the actual movement trajectory of the sorted cargo according to the monitoring image sequence within a preset historical period; and determine the instability value of the sorted cargo according to the actual movement trajectory.

[0027] For more information about the acquisition module 110, the first determination module 120, the second determination module 130 and the feedback module, see Figure 2-Figure 4 Related description.

[0028] It should be understood that Figure 1 The system and its modules shown can be implemented in various ways. For example, in some embodiments, the sorting machine sorting control system 100 can be implemented by a processor. The processor can process data and / or information obtained from other devices or system components. The processor can execute program instructions based on these data, information and / or processing results to perform one or more functions described in the embodiments of this specification. As an example only, the processor may include a central processing unit (CPU), an application specific integrated circuit (ASIC), etc. or any combination of the above. In some embodiments, the processor may include multiple modules, and different modules can be used to execute different program instructions respectively.

[0029] It should be noted that the above description of the sorting control system 100 and its modules is only for the convenience of description and cannot limit the present specification to the scope of the embodiments. It is understandable that for those skilled in the art, after understanding the principle of the system, it is possible to arbitrarily combine the modules or form a subsystem to connect with other modules without deviating from this principle. In some embodiments, Figure 1 The acquisition module, the first determination module, and the second determination module disclosed in the specification may be different modules in a system, or one module may realize the functions of two or more modules. For example, each module may share a storage module, or each module may have its own storage module. Such variations are within the protection scope of this specification.

[0030] Figure 2 is an exemplary flow chart of a sorting control method for a sorting machine according to some embodiments of this specification. Figure 2 As shown, the process 200 includes the following steps. In some embodiments, the process 200 can be executed by the sorting control system 100 of the sorting machine.

[0031] The sorting machine can sort the items according to pre-set instructions and transport the sorted items to a designated location. In some embodiments, the sorting machine can be widely used in logistics, manufacturing, warehousing and other fields. In some embodiments, the sorting machine can include a pendulum sorting machine, a conveyor belt sorting machine, a wheel sorting machine, a robotic arm sorting machine, a slide sorting machine, a digital sorting machine, etc. The relevant description of the above-mentioned sorting machine is for illustrative purposes only and is not intended to limit the scope of this specification. In the embodiments of this specification, a pendulum sorting machine is taken as an example for explanation.

[0032] Step 210, obtaining cargo characteristics of the target cargo and conveying characteristics of the conveyor belt.

[0033] The target cargo refers to cargo that is currently on the conveyor belt and whose distance to the balance belt is less than a preset distance threshold.

[0034] Goods refer to items to be sorted and transported by the sorting machine. In some embodiments, the goods may include express parcels, airport luggage, goods in e-commerce logistics, and various objects in garbage classification.

[0035] Conveyor belt refers to the equipment used to transport materials or items in the balance wheel sorter. Balance belt refers to the equipment that performs sorting in the balance wheel sorter. When the goods are transported on the conveyor belt, they will continue to approach the balance belt until the goods reach the balance belt. Then, they will be slid out of the current conveyor belt in a specific direction through the rollers of the balance belt to complete the sorting.

[0036] The preset distance threshold may be preset based on experience. In some embodiments, the preset distance threshold may include 1m, 1.5m, 2m, 2.5m, 3m, etc.

[0037] Cargo characteristics refer to characteristics related to the target cargo. In some embodiments, the cargo characteristics may include the height, weight, and placement posture of the target cargo. The placement posture refers to the minimum deviation angle of the target cargo along the transport direction of the conveyor belt. For example, if the target cargo has two deviation angles a and deviation angle b along the transport direction of the conveyor belt, and the deviation angle a is greater than the deviation angle b, then the smallest deviation angle b can be selected as the placement posture. Among them, when the central axis of the target cargo is not parallel to the transport direction of the conveyor belt, it can be regarded as the target cargo deviating from the transport direction of the conveyor belt. The deviation angle refers to the angle between the central axis of the target cargo and the transport direction of the conveyor belt.

[0038] In some embodiments, the acquisition module 110 can acquire cargo features based on cargo images captured when the target cargo passes through the inspection stage. The inspection stage refers to the stage where the cargo is inspected before sorting. For example, the inspection stage can perform security checks on the cargo. During the inspection stage, an image acquisition device can be set to acquire cargo images.

[0039] In some embodiments, the acquisition module 110 may perform image recognition on the cargo image through a preset algorithm to obtain cargo features. For example, the preset algorithm may include an image recognition algorithm.

[0040] In some embodiments, the acquisition module 110 may also acquire cargo images based on an image acquisition device disposed around the conveyor belt. The image acquisition device may include a camera, a laser radar, an infrared sensor, etc. In some embodiments, the acquisition module 110 may acquire multiple cargo images of the target cargo at multiple positions on the conveyor belt during the conveying process, and determine multiple placement postures based on the multiple cargo images.

[0041] When the goods are running on the conveyor belt, the goods may be offset, so the placement posture of the goods during the conveying process may change accordingly. It can be understood that the placement posture determined based on the goods image in the inspection stage is the initial placement posture, and the placement posture determined based on the goods image in the conveying stage is the placement posture during the conveying process.

[0042] The conveying characteristics refer to characteristics related to the conveyor belt. In some embodiments, the conveying characteristics include the conveying speed and cargo interval of the conveyor belt. The conveying speed refers to the distance conveyed by the conveyor belt in a unit time. The cargo interval refers to the distance between two goods that are continuously conveyed. In some embodiments, the acquisition module 110 can directly acquire the conveying characteristics by setting the conveyor belt parameters.

[0043] Step 220, based on the cargo characteristics and the conveying characteristics, determine the estimated initial velocity and the estimated attitude angle of the target cargo when it arrives at the balance belt.

[0044] The estimated initial velocity refers to the speed of the target goods when being sorted. In some embodiments, the first determination module 120 can determine the estimated initial velocity according to the conveying speed in the conveying feature. For example, the estimated initial velocity can be the conveying speed of the conveyor belt.

[0045] The estimated attitude angle refers to the deviation angle of the target cargo when it reaches the balance belt.

[0046] In some embodiments, the first determination module 120 can obtain the cargo characteristics (e.g., placement posture) of the target cargo at multiple positions on the conveyor belt, and determine the estimated posture angle according to the multiple cargo characteristics. Among them, the multiple positions on the conveyor belt can be positions on the conveyor belt that are spaced at a fixed distance. In some embodiments, the interval between two adjacent positions in the multiple positions can be a preset interval threshold. The preset interval threshold can be set artificially based on experience. In some embodiments, the preset interval threshold can include 1m, 1.5m, 2m, 2.5m, 3m, etc.

[0047] In some embodiments, the first determination module 120 can determine the estimated attitude angle according to the change of the placement attitude of the multiple cargo features with the transmission distance. The transmission distance is the distance that the target cargo moves along the conveyor belt. The first determination module 120 can determine the change of the deviation angle per unit distance according to the change of the placement attitude of the target cargo with the transmission distance to estimate the deviation angle of the target cargo when it arrives at the balance belt.

[0048] In some embodiments, the first determination module 120 can obtain the relationship between the placement posture and distance of the target cargo through polynomial fitting based on the distances of multiple position points from the balance belt and the corresponding placement postures, and then estimate the deviation angle of the target cargo when it reaches the balance belt based on the relationship between the placement posture and the distance.

[0049] As an example only, the first determination module 120 can obtain the cargo characteristics of the target cargo at a distance of 10m, 8m, 6m, and 4m from the balance belt, respectively, and the corresponding placement postures are 20°, 18°, 16°, and 14°, respectively. The relationship between the placement posture and the distance can be determined through polynomial fitting, that is, the placement posture changes by 1° every 1m, thereby calculating the deviation angle of the target cargo when it arrives at the balance belt, that is, the estimated posture angle is 12°.

[0050] Step 230, determining the roller parameter matrix of the balance belt based on the cargo characteristics, estimated initial velocity, estimated attitude angle and target sorting direction of the target cargo.

[0051] The target sorting direction refers to the direction in which the target goods need to be sorted. The target sorting direction can be expressed as an angle with the transmission direction of the conveyor belt.

[0052] The roller parameter matrix refers to a matrix composed of roller parameters of multiple rollers of the balance belt. The roller parameters refer to parameters related to the rollers. In some embodiments, the roller parameters may include roller angles. The roller angle refers to the angle between the roller and the workbench (e.g., a conveyor belt). The change of the roller angle can change the sorting direction of the balance belt. The sorting direction of the balance belt refers to the direction in which the goods are sorted when passing through the balance belt.

[0053] In some embodiments, the roller parameter matrix of the balance wheel belt can be set by the staff according to actual conditions and experience, and the second determination module 130 can directly obtain the roller parameter matrix set by the staff.

[0054] In some embodiments, the second determination module 130 can determine the sorting abnormality probability of the target cargo based on the cargo characteristics, estimated initial velocity, estimated attitude angle and target sorting direction; in response to the sorting abnormality probability being less than a preset threshold, determine the roller parameter matrix based on the cargo characteristics, estimated initial velocity, estimated attitude angle and target sorting direction.

[0055] The sorting anomaly probability refers to the probability of anomalies (i.e., sorting anomalies) occurring when the target goods are sorted. In some embodiments, the sorting anomaly may include the target goods being stuck on the balance belt or the target goods being thrown in an undesirable direction when passing through the balance belt. An undesirable direction is a direction other than the target sorting direction.

[0056] In some embodiments, for each sorting direction of the balance wheel belt, the second determination module 130 can cluster the historical data of the sorting direction based on the cargo characteristics, estimated initial velocity, and estimated attitude angle of the historical sorted cargo, and statistically calculate the probability of sorting abnormality for each category in the clustering result as the sorting abnormality probability of the category. The historical data is the relevant data of the sorting of historical sorted cargo in the historical time. The historical data may include cargo characteristics, estimated initial velocity, estimated attitude angle, and corresponding conveying characteristics of historical sorted cargo when it was sorted. The clustering algorithm may include, but is not limited to, K-Means clustering and / or density-based clustering method (DBSCAN), etc.

[0057] In some embodiments, the second determination module 130 may select the sorting anomaly probability of the category closest to the target cargo as the sorting anomaly probability of the target cargo.

[0058] In some embodiments, the second determination module 130 can construct a sorting feature vector based on cargo characteristics, estimated initial velocity, estimated attitude angle, and target sorting direction; and match the sorting feature vector in a preset database to determine a roller parameter matrix. The preset database includes multiple reference sorting feature vectors and corresponding reference roller parameter matrices.

[0059] In some embodiments, the second determination module 130 can construct at least one reference sorting feature vector in the same manner as constructing the sorting feature vector based on the historical cargo characteristics, historical estimated initial velocity, historical estimated attitude angle, and historical target sorting direction of the historical target cargo in the historical data, and use the preferred roller parameter matrix as the reference roller parameter matrix of the reference sorting feature vector. In some embodiments, the second determination module 130 can cluster at least one reference sorting feature vector to obtain at least one clustering result; determine the target clustering result from at least one clustering result, and determine the reference roller parameter matrix corresponding to the target clustering result based on the reference roller parameter matrices of the multiple reference sorting feature vectors included in the target clustering result. Each reference sorting feature vector in the target clustering result corresponds to the reference roller parameter matrix corresponding to the target clustering result. Based on the above method, multiple reference sorting feature vectors and corresponding reference roller parameter matrices can be determined and stored in a preset database.

[0060] In some embodiments, the second determination module 130 may determine a clustering result in which no sorting anomaly occurs in multiple sortings corresponding to multiple reference sorting feature vectors in at least one clustering result as a target clustering result.

[0061] In some embodiments, the second determination module 130 may use the mean value of the roller parameter matrices corresponding to the multiple reference sorting feature vectors included in the target clustering result as the reference roller parameter matrix corresponding to the target clustering result.

[0062] In some embodiments, the second determination module 130 may determine the reference roller parameter matrix corresponding to the reference sorting feature vector closest to the sorting feature vector (i.e., the reference sorting feature vector with the highest similarity to the sorting feature vector) in the preset database as the roller parameter matrix.

[0063] The preset threshold is a threshold related to the probability of sorting anomaly. The preset threshold can be set based on human experience.

[0064] In some embodiments, the preset threshold value may be related to the cargo interval. In some embodiments, the preset threshold value may be positively correlated to the cargo interval. For example, the smaller the cargo interval, the smaller the preset threshold value.

[0065] In the embodiments of this specification, the smaller the interval between goods, the greater the chain reaction when a certain goods is sorted abnormally. At this time, the preset threshold needs to be appropriately lowered to control the risk of abnormality during sorting. Therefore, associating the preset threshold with the interval between goods can reduce the risk of abnormality during sorting and avoid various problems that may occur during the sorting and transportation of goods as much as possible.

[0066] In some embodiments, the second determination module 130 may also determine a preset threshold value according to the instability distribution and the cargo interval of the conveyor belt.

[0067] In some embodiments, the second determination module 130 may use the probability mean of multiple unstable state combinations in the unstable distribution as the initial threshold. In some embodiments, the second determination module 130 may determine the adjustment ratio based on the cargo interval and the standard cargo interval of the conveyor belt; in response to the cargo interval of the conveyor belt being not greater than the standard cargo interval, the initial threshold is adjusted according to the adjustment ratio to obtain a preset threshold. In some embodiments, the second determination module 130 may determine the difference between the cargo interval and the standard cargo interval of the conveyor belt, and determine the ratio of the difference to the cargo interval of the conveyor belt as the adjustment ratio. Among them, the standard cargo interval can be preset based on experience. For more information about unstable distribution and unstable state combinations, please refer to Figure 4 and related instructions.

[0068] In some embodiments, in response to the cargo interval of the conveyor belt being greater than the standard cargo interval, the second determination module 130 may not adjust the initial threshold value, and directly determine the initial threshold value as the preset threshold value.

[0069] In the embodiments of the present specification, by determining the preset threshold value based on the instability distribution and the cargo interval, the possibility of cargo instability can be avoided as much as possible, thereby avoiding various problems that may occur during the cargo sorting and transportation process.

[0070] In some embodiments, in response to the sorting abnormality probability being less than a preset threshold, the second determination module 130 may determine the roller parameter matrix through a preset database based on cargo characteristics, estimated initial velocity, estimated attitude angle, and target sorting direction.

[0071] For how to determine the roller parameter matrix through the preset database, please refer to the above description.

[0072] In the embodiment of the present specification, by determining the probability of sorting anomalies to determine the roller parameter matrix, the situation where the probability of sorting anomalies is too high can be eliminated, thereby reducing the risk of anomalies occurring during sorting.

[0073] In some embodiments, in response to the sorting abnormality probability being greater than a preset threshold, the second determination module 130 may also generate a plurality of candidate roller parameter matrices according to the target sorting direction; for each candidate roller parameter matrix, based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle and the candidate roller parameter matrix, the stability of the target cargo under the candidate roller parameter matrix is ​​determined through a stability prediction model; the roller parameter matrix is ​​determined based on the stabilities corresponding to each of the plurality of candidate roller parameter matrices.

[0074] The candidate roller parameter matrix is ​​a roller parameter matrix that is a candidate.

[0075] In some embodiments, the second determination module 130 can randomly generate a large number of candidate roller parameter matrices that deliver the target goods to the target sorting direction based on the target sorting direction. For example, the second determination module 130 can determine the parameter range of each roller parameter in the roller parameter matrix based on the target sorting direction; based on the parameter range of each roller parameter in the roller parameter matrix, randomly generate multiple values ​​of each roller parameter in the roller parameter matrix, and combine them in any feasible combination to obtain multiple candidate roller parameter matrices. In some embodiments, the second determination module 130 can determine the parameter range of each roller parameter corresponding to the target sorting direction based on historical data. For example, the second determination module 130 can determine the parameter range of each roller parameter corresponding to the target sorting direction based on the range of each roller parameter corresponding to the target sorting direction in the historical data.

[0076] Stability is a parameter that evaluates the possibility that the target goods are sorted along the target sorting direction when they are sorted under the candidate roller parameter matrix. The lower the stability, the more likely the target goods are to deviate from the target sorting direction.

[0077] In some embodiments, the stability prediction model is a machine learning model. In some embodiments, the stability prediction model can be a convolutional neural network (CNN) model and a deep neural network (DNN) model. The stability prediction model can also be a machine learning model of other structures, such as a neural network model, a recurrent neural network model, etc.

[0078] In some embodiments, the input of the stability prediction model may include cargo characteristics, estimated initial velocity, estimated attitude angle, and candidate roller parameter matrix, and the output may include the stability of the target cargo under the candidate roller parameter matrix.

[0079] In some embodiments, the input to the stability prediction model may also include roller distribution characteristics.

[0080] The roller distribution feature refers to a feature that characterizes the distribution of data related to the roller. In some embodiments, the roller distribution feature may include the width of the roller and the gap between the rollers.

[0081] The width of the roller refers to the length of the roller in the direction of the rotation axis. The rotation axis of the roller is the central axis of the roller when it rotates. The width of the roller determines the contact area between the roller and the goods. A larger contact area can increase the stability of the goods during the sorting process.

[0082] The gap between rollers refers to the space between the roller and the bearing ring. The gap between rollers determines the frequency of alternating rolling and sliding of the goods, so the gap between rollers will affect the stability.

[0083] In the embodiments of the present specification, the roller distribution characteristics are further used as the input of the model, which can make the stability of the model output more accurate, reduce the risk of goods tilting or falling, and more easily avoid problems in the sorting process.

[0084] In some embodiments, the second determination module 130 may obtain a plurality of first training samples with a first label to form a first training sample set, and obtain a stability prediction model through training with the first training sample set. The first training sample may be obtained from historical data. In some embodiments, the first training sample may be obtained based on historical data of a plurality of sample conveyor belts.

[0085] In some embodiments, the first training sample may include sample cargo features of the sample cargo, sample estimated initial velocity of the sample cargo on the sample conveyor belt, sample estimated attitude angle of the sample cargo on the sample conveyor belt, and a sample roller parameter matrix. The first label may be the stability corresponding to the sample roller parameter matrix. The first label may be determined by the system or by human annotation.

[0086] In some embodiments, the second determination module 130 may input the first training sample set into the initial stability prediction model to perform multiple rounds of iterations. Each round of iteration includes: selecting one or more first training samples from the first training sample set, inputting one or more first training samples into the initial stability prediction model, and obtaining the model prediction output corresponding to the one or more first training samples; according to the model prediction output corresponding to the one or more first training samples, and the first label of the one or more first training samples, substitute into the formula of the predefined loss function to calculate the value of the loss function; according to the value of the loss function, reversely update the model parameters in the initial stability prediction model. When the iteration end condition is met, the iteration ends and a trained fault prediction model is obtained. Among them, the iteration end condition may be that the loss function converges, the number of iterations reaches a threshold, etc.

[0087] In some embodiments, the number of training samples corresponding to each sample conveyor belt is at least greater than a preset number threshold. The preset number threshold may be related to the standard deviation of multiple historical cargo characteristics. The larger the standard deviation of the historical cargo characteristics, the larger the preset number threshold may be.

[0088] The historical cargo characteristics refer to the cargo characteristics of the sorted cargo within the historical period. Each sorted cargo corresponds to a historical cargo characteristic. In some embodiments, the mean of the standard deviations of the parameters in the multiple historical cargo characteristics can be determined as the standard deviation of the multiple historical cargo characteristics. For example, when the historical cargo characteristics include two parameters, the standard deviations of the two parameters can be determined respectively, and then the mean of the standard deviations of the two parameters can be determined as the standard deviation of the multiple historical cargo characteristics.

[0089] In the embodiment of this specification, the larger the standard deviation of the historical cargo characteristics, the more types of cargo are sorted in the sorting area, and the greater the difference in characteristics between the cargoes. At this time, it is necessary to appropriately increase the number of samples collected on each conveyor belt to ensure that the samples can cover more sorting situations.

[0090] In some embodiments, the stability prediction model can also be a machine learning model with a custom structure as described below. Figure 3 As shown, the stability prediction model 310 may include a prediction layer 320 and an evaluation layer 330. The prediction layer 320 and the evaluation layer 330 may be obtained by separate training.

[0091] In some embodiments, the prediction layer may be a convolutional neural network (CNN) model.

[0092] like Figure 3 As shown, in some embodiments, the input of the prediction layer 320 may include cargo features 321 , estimated initial velocity 322 , estimated attitude angle 323 , and candidate roller parameter matrix 324 , and the output of the prediction layer 320 may include a motion trajectory 340 .

[0093] The motion trajectory includes the predicted position and predicted speed of the target cargo at multiple time points. The predicted speed includes the speed magnitude and speed direction. In some embodiments, the second determination module 130 can establish a coordinate system with the center of the balance belt as the origin and the directions parallel to the sides of the balance belt on the horizontal plane as the x-axis and y-axis to express the predicted position in coordinate form.

[0094] In some embodiments, the prediction layer 320 may obtain a plurality of second training samples with second labels to form a second training sample set, and may be trained by the second training sample set. In some embodiments, the prediction layer 320 may be obtained by training similar to the training process of the stability prediction model described above.

[0095] In some embodiments, the second training sample may include sample cargo features of the sample cargo, sample estimated initial velocity of the sample cargo on the sample conveyor belt, sample estimated attitude angle of the sample cargo on the sample conveyor belt, and sample candidate roller parameter matrix. The second label may be the motion trajectory corresponding to the second training sample. In some embodiments, the second label may be determined by the actual motion trajectory of the sample cargo corresponding to the second training sample in the historical data after it subsequently enters the balance belt.

[0096] In some embodiments, the evaluation layer can be a deep neural network (DNN) model.

[0097] like Figure 3 As shown, in some embodiments, the input of the evaluation layer 330 may include cargo features 321 and motion trajectory 340, and the output of the evaluation layer 330 may include stability 350 corresponding to the motion trajectory 340, that is, the stability 350 of the target cargo under the candidate roller parameter matrix.

[0098] In some embodiments, the evaluation layer 330 may obtain a plurality of third training samples with third labels to form a third training sample set, and may be trained by the third training sample set. In some embodiments, the evaluation layer 330 may be obtained by training similar to the training process of the stability prediction model described above.

[0099] In some embodiments, the third training sample may include sample cargo features and sample motion trajectories of the sample cargo. The third label may be the stability of the sample motion trajectory corresponding to the third training sample. In some embodiments, the second determination module 130 may cluster the third training samples in the third training sample set, and for each type of training samples obtained by clustering, the ratio of the total number of training samples of this type to the number of deviated samples in this type of training samples is used as the third label corresponding to this type of training sample. Among them, the deviated samples are training samples in which the cargo deviates from the predetermined trajectory during subsequent sorting. The predetermined trajectory may be a preset standard trajectory or an average trajectory of a large number of normally sorted cargoes. Normal sorting refers to sorting without falling or other abnormalities. Deviating from the predetermined trajectory refers to the situation where the overlap value of the actual trajectory with the predetermined trajectory is less than the set trajectory threshold. The trajectory threshold can be set through human experience.

[0100] In some embodiments, the second determination module 130 can determine the overlap value of the actual trajectory and the predetermined trajectory by a variety of methods. For example, the position and speed of each time point in the actual trajectory and the predetermined trajectory are compared separately, and if they overlap, they are recorded as 1, and if they do not overlap, they are recorded as a value of 0 to 1, and finally the values ​​of all time points are added to obtain the overlap value.

[0101] In some embodiments, the second determination module 130 may select a candidate roller parameter matrix with the highest stability as the roller parameter matrix of the balance belt.

[0102] In the embodiments of this specification, the stability of the target cargo under the candidate roller parameter matrix is ​​determined by the stability prediction model, and the self-learning ability of the machine learning model can be used to find patterns from a large amount of historical data to obtain cargo characteristics, estimated initial velocity, estimated attitude angle, and the relationship between the candidate roller parameter matrix and stability, thereby improving the accuracy and efficiency of determining stability. By determining the stability corresponding to the candidate roller parameter matrix to determine the roller parameter matrix, the possibility of the target cargo being sorted in the target sorting direction can be increased, and the target cargo can be prevented from deviating from the target sorting direction as much as possible.

[0103] In the embodiments of this specification, the roller parameter matrix of the balance wheel belt is determined by cargo characteristics, estimated initial velocity, estimated attitude angle, and target sorting direction, and the relevant parameters of the balance wheel sorting machine can be evaluated and adjusted in advance, so as to avoid various problems that may occur during the cargo sorting and transportation process as much as possible, achieve accurate and efficient cargo sorting, and ensure that the cargo is sorted and transported to the target location.

[0104] It should be noted that the above description of the process 200 is only for example and illustration, and does not limit the scope of application of this specification. For those skilled in the art, various modifications and changes can be made to the process 200 under the guidance of this specification. However, these modifications and changes are still within the scope of this specification.

[0105] Figure 4 It is an exemplary schematic diagram of a sorting control method for a sorting machine according to some embodiments of the present specification.

[0106] In some embodiments, Figure 4 As shown, the feedback module can determine the instability distribution 420 by calculating the instability values ​​410 of multiple sorted goods within a preset historical period; determine whether the instability distribution 420 meets the preset condition 430; in response to the instability distribution 420 meeting the preset condition 430, generate and send a fault warning 440, and / or adjust the running speed 450 of the conveyor belt.

[0107] The preset historical period is a period of a specific length of time before the current moment. The preset historical period can be set based on experience. In some embodiments, the preset historical period can be the previous hour, the previous 3 hours, the previous 10 hours, the previous day, the previous 3 days, the previous 5 days, etc.

[0108] The instability value is the probability of cargo instability. Instability refers to the situation where cargo is not sorted and transported smoothly and safely. Instability can include cargo shaking, tilting, etc.

[0109] In some embodiments, the feedback module can determine the instability value through a preset table based on the cargo characteristics and historical transportation characteristics of a plurality of sorted cargoes within a preset historical period. The preset table is a table including cargo characteristics, transportation characteristics and corresponding instability values. In some embodiments, the preset table can be constructed through human experience.

[0110] In some embodiments, for each sorted cargo within a preset historical period, the feedback module may search a preset table for an instability value corresponding to the cargo characteristics and the transport characteristics of the sorted cargo as the instability value of the sorted cargo.

[0111] In some embodiments, for each sorted cargo, the feedback module may determine the actual movement trajectory of the sorted cargo according to a monitoring image sequence within a preset historical period; and determine the instability value of the sorted cargo according to the actual movement trajectory.

[0112] The monitoring image sequence refers to a sequence of monitoring images of sorted goods. In some embodiments, the feedback module can obtain monitoring images through an image acquisition device (e.g., a camera) in the room where the sorter is located. In some embodiments, the feedback module can obtain monitoring images within a preset historical period to form a monitoring image sequence.

[0113] The actual motion trajectory refers to the actual motion trajectory of the sorted goods during sorting. In some embodiments, the feedback module can identify the monitoring image sequence through an image recognition algorithm to determine the actual motion trajectory. In some embodiments, the image recognition algorithm can include a frame difference method, an optical flow method, and a deep learning method.

[0114] In some embodiments, the feedback module can select multiple trajectory points on the actual motion trajectory, calculate the derivatives of the actual motion trajectory at the multiple trajectory points, and use the standard deviation of the multiple derivatives as the instability value of the sorted goods. In some embodiments, the feedback module can establish a coordinate system with the center of the balance belt as the origin, and the direction parallel to the edge of the balance belt on the horizontal plane as the x-axis and y-axis, and express the coordinate positions of the multiple trajectory points in the form of coordinates. For each trajectory point, the feedback module can calculate the derivative of the actual motion trajectory at the current trajectory point based on the coordinates of the current trajectory point and the coordinates of the previous trajectory point.

[0115] In some embodiments, the feedback module can calculate the derivative of the actual motion trajectory at the current trajectory point by formula (1):

[0116] f t =y t-1 -y t / (x t-1 -x t ) (1)

[0117] Among them, t is the sequence number of the trajectory point, f represents the derivative of the actual motion trajectory at the current trajectory point, yt-1 represents the y-axis coordinate of the previous trajectory point, and y t Indicates the y-axis coordinate of the current trajectory point, x t-1 represents the x-axis coordinate of the previous trajectory point, and x represents the x-axis coordinate of the current trajectory point.

[0118] In the embodiment of this specification, the standard deviation of the derivative on the motion trajectory can reflect the smoothness of the motion trajectory. The less smooth the current motion trajectory is, the more times the goods become unstable during sorting. Therefore, determining the instability value through the actual motion trajectory can make the obtained instability value more accurate.

[0119] The unstable distribution is the distribution of unstable values ​​of multiple sorted goods. The unstable distribution can be a polynomial distribution.

[0120] The preset condition is a condition for determining an unstable distribution. In some embodiments, the preset condition may include that the probability mean of a plurality of unstable state combinations in the unstable distribution is greater than an instability threshold. The instability threshold may be set based on experience.

[0121] The instability state combination indicates a possible situation of whether the multiple sorted goods are instability. Different instability state combinations indicate situations in which different goods among the multiple sorted goods are instability. In some embodiments, the feedback module may arrange and combine the two states of whether each of the multiple sorted goods is instability to obtain multiple instability state combinations.

[0122] Just as an example, if there are three sorted goods A, B, and C in the preset historical period, and the corresponding instability values ​​are pA, pB, and pC respectively, then eight unstable state combinations can be determined, including: X=(0, 0, 0), indicating that all sorted goods are not unstable; X2=(0, 0, 1), indicating that goods C is unstable; X3=(0, 1, 0), indicating that goods B is unstable; X4=(0, 1, 1) indicates that goods B and goods C are unstable; ...

[0123] In some embodiments, the feedback module can obtain multiple unstable state combinations and their probabilities based on the unstable distribution. The probability of the unstable state combination represents the probability that multiple sorted goods in the unstable distribution are in the unstable state combination. The probability of each unstable state combination can be expressed as formula (2):

[0124]

[0125] Where X represents the unstable state combination, X = (x1, x2, x3, ..., x n ), p i Represents x i The instability value, x iIndicates whether the sorted item i is unstable, x i =1 means that the sorted cargo i is unstable, x i =0 means that the sorted cargo i is not unstable.

[0126] Continuing with the above example, for the unstable state combination X4 = (0, 1, 1), the probability P(X4) of X4 can be expressed as formula (3):

[0127] P(X4)=(1-p A )×p B ×p C (3)

[0128] In some embodiments, in response to the instability distribution meeting a preset condition, the feedback module may generate and send a fault warning.

[0129] Fault warning is a warning of possible faults that may occur within a preset historical period. The fault warning may include the instability values, instability distribution, multiple instability state combinations and their corresponding faults of the multiple sorted goods determined above. In some embodiments, the generated fault warning may be sent to a device that interacts with the user, such as a user terminal. After receiving the fault warning, the user may process it according to the content of the fault warning to solve the fault.

[0130] In some embodiments, in response to the instability distribution satisfying a preset condition, the feedback module may further adjust the conveying speed of the conveyor belt.

[0131] For more information on conveying speed, see Figure 2 and related instructions.

[0132] In some embodiments, the feedback module can determine a speed adjustment ratio, and reduce the conveying speed of the conveyor belt according to the speed adjustment ratio. In some embodiments, the feedback module can determine a difference between a probability mean of a plurality of unstable state combinations and an instability threshold, and use the ratio of the difference to the instability threshold as the speed adjustment ratio.

[0133] In the embodiments of the present specification, by determining the instability distribution to generate a fault warning or adjust the running speed of the conveyor belt, the probability of a possible fault can be judged based on the situation of the sorted goods in a preset historical period, so that the possible fault can be processed to solve the fault, or the conveying speed of the conveyor belt can be reduced to avoid the recurrence of the fault as much as possible.

[0134] One or more embodiments of the present specification further provide a sorting control device for a sorting machine, the device comprising at least one processor and at least one memory; the at least one memory is used to store computer instructions; the at least one processor is used to execute at least part of the computer instructions to implement a sorting control method for a sorting machine as described in any embodiment of the present specification.

[0135] One or more embodiments of the present specification further provide a computer-readable storage medium, wherein the storage medium stores computer instructions, and is characterized in that after the computer reads the computer instructions in the storage medium, the computer executes the sorting control method of the sorting machine as described in any embodiment of the present specification.

[0136] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of this specification. Although not explicitly stated here, 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, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of this specification.

[0137] At the same time, this specification uses specific words to describe the embodiments of this specification. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of this specification. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures or characteristics in one or more embodiments of this specification can be appropriately combined.

[0138] In addition, unless explicitly stated in the claims, the order of the processing elements and sequences described in this specification, the use of alphanumeric characters, or the use of other names are not intended to limit the order of the processes and methods of this specification. Although the above disclosure discusses some invention embodiments that are currently considered useful through various examples, it should be understood that such details are only for illustrative purposes, and the attached claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the essence and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

[0139] Similarly, it should be noted that in order to simplify the description disclosed in this specification and thus help understand one or more embodiments of the invention, in the above description of the embodiments of this specification, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this disclosure method does not mean that the features required by the subject matter of this specification are more than the features mentioned in the claims. In fact, the features of the embodiments are less than all the features of the single embodiment disclosed above.

[0140] In some embodiments, numbers describing the number of components and attributes are used. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise specified, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may change according to the required features of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of this specification are approximate values, in specific embodiments, the setting of such numerical values ​​is as accurate as possible within the feasible range.

[0141] Each patent, patent application, patent application publication, and other materials, such as articles, books, specifications, publications, documents, etc., cited in this specification are hereby incorporated by reference in their entirety. Except for application history documents that are inconsistent with or conflicting with the contents of this specification, documents that limit the broadest scope of the claims of this specification (currently or later attached to this specification) are also excluded. It should be noted that if the descriptions, definitions, and / or use of terms in the materials attached to this specification are inconsistent or conflicting with the contents described in this specification, the descriptions, definitions, and / or use of terms in this specification shall prevail.

[0142] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, as an example and not a limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A sorting control method for a sorting machine, characterized in that: include: Acquire cargo characteristics of target cargo and conveying characteristics of the conveyor belt, wherein the target cargo is cargo that is currently on the conveyor belt and whose distance from the swing belt is less than a preset distance threshold, the cargo characteristics include the height, weight and placement posture of the target cargo, and the conveying characteristics include the conveying speed of the conveyor belt and cargo interval; Based on the cargo characteristics and the conveying characteristics, determining an estimated initial velocity and an estimated attitude angle of the target cargo when it arrives at the balance belt; Determining a sorting abnormality probability of the target cargo based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle, and the target sorting direction of the target cargo; In response to the sorting abnormality probability being less than a preset threshold, determining a roller parameter matrix of the balance belt based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle, and the target sorting direction; Determine the instability distribution by calculating the instability values ​​of multiple sorted goods within a preset historical period; In response to the instability distribution satisfying a preset condition, a fault warning is generated and sent, and / or the conveying speed of the conveyor belt is adjusted.

2. The method according to claim 1, characterized in that The method further comprises: In response to the sorting abnormality probability being greater than the preset threshold: According to the target sorting direction, generating a plurality of candidate roller parameter matrices; For each candidate roller parameter matrix, based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle and the candidate roller parameter matrix, the stability of the target cargo under the candidate roller parameter matrix is ​​determined by a stability prediction model, wherein the stability prediction model is a machine learning model; The roller parameter matrix is ​​determined based on the stabilities corresponding to each of the plurality of candidate roller parameter matrices.

3. The method according to claim 1, characterized in that The calculating of the instability values ​​of a plurality of sorted goods within a preset historical period includes: For each of the sorted goods, Determining the actual movement trajectory of the sorted goods according to the monitoring image sequence within the preset historical period; According to the actual motion trajectory, an instability value of the sorted goods is determined.

4. A sorting control system for a sorting machine, characterized in that: It includes an acquisition module, a first determination module and a second determination module; The acquisition module is configured to acquire cargo characteristics of target cargo and conveying characteristics of the conveyor belt, wherein the target cargo is cargo that is currently on the conveyor belt and whose distance from the balance belt is less than a preset distance threshold, the cargo characteristics include the height, weight and placement posture of the target cargo, and the conveying characteristics include the conveying speed of the conveyor belt and cargo interval; The first determination module is configured to determine an estimated initial velocity and an estimated attitude angle of the target cargo when it arrives at the balance belt based on the cargo characteristics and the transport characteristics; The second determination module is configured to determine the sorting abnormality probability of the target cargo based on the cargo characteristics of the target cargo, the estimated initial velocity, the estimated attitude angle, and the target sorting direction of the target cargo; In response to the sorting abnormality probability being less than a preset threshold, determining a roller parameter matrix of the balance belt based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle, and the target sorting direction; Determine the instability distribution by calculating the instability values ​​of multiple sorted goods within a preset historical period; In response to the instability distribution satisfying a preset condition, a fault warning is generated and sent, and / or the conveying speed of the conveyor belt is adjusted.

5. The system according to claim 4, characterized in that The second determining module is further configured to: In response to the sorting abnormality probability being greater than the preset threshold: According to the target sorting direction, generating a plurality of candidate roller parameter matrices; For each candidate roller parameter matrix, based on the cargo characteristics, the estimated initial velocity, the estimated attitude angle and the candidate roller parameter matrix, the stability of the target cargo under the candidate roller parameter matrix is ​​determined by a stability prediction model, wherein the stability prediction model is a machine learning model; The roller parameter matrix is ​​determined based on the stabilities corresponding to each of the plurality of candidate roller parameter matrices.

6. A sorting control device for a sorting machine, comprising a processor, wherein the processor is used to execute the sorting control method for a sorting machine according to any one of claims 1 to 3.

7. A computer-readable storage medium, wherein the storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes the sorting control method for a sorting machine as claimed in any one of claims 1 to 3.

Images