A multi-source demand-oriented logistics task intelligent allocation method and system
By using computing power boxes to collect data and interact with and fuse anomaly feature anchor vectors in logistics tasks, an intelligent allocation scheme is generated, which solves the problem of uneven allocation of logistics tasks and improves resource utilization and task response efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BRINGSPRING SCIENCE & TECHNOLOGY CO LTD
- Filing Date
- 2025-06-25
- Publication Date
- 2026-06-16
AI Technical Summary
The existing logistics task allocation method lacks the ability to uniformly analyze and intelligently process multi-source demand data, resulting in uneven task allocation, low resource utilization, and difficulty in achieving efficient and refined scheduling.
Data is collected by M computing boxes corresponding to M logistics devices distributed in the target area, anomaly feature anchor vectors are extracted, attention interaction is performed to generate logistics task work orders, and intelligent allocation is performed in combination with logistics resource distribution information.
It achieved efficient matching of tasks and resources, improved the utilization rate of logistical resources, and optimized task response efficiency and resource utilization level.
Smart Images
Figure CN120706815B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of task allocation technology, and specifically to a method and system for intelligent allocation of logistics tasks oriented towards multi-source needs. Background Technology
[0002] In the field of logistics management, with the increasing scale and complexity of facilities such as smart parks, intelligent buildings, hospitals, schools, and enterprises, daily logistics services encompass multiple aspects, including equipment operation and maintenance, material management, fire safety, security patrols, and environmental protection, exhibiting significant characteristics of diverse needs, complex tasks, and dynamic resources. Existing logistics task allocation methods largely rely on manual management or preset fixed rules, lacking the ability to uniformly analyze and intelligently process multi-source logistics demand data. This makes it difficult to promptly identify the priority and urgency of different tasks, and hinders the efficient, refined scheduling and rational allocation of logistics resources. Overall response efficiency and resource utilization levels need improvement. Summary of the Invention
[0003] This application provides a method and system for intelligent allocation of logistics tasks to meet multi-source needs, which solves the technical problems of uneven allocation of logistics tasks and low resource utilization in the prior art.
[0004] The first aspect of this application provides a method for intelligent allocation of logistics tasks oriented towards multi-source needs, the method comprising:
[0005] The system interacts with M logistics devices distributed across a target area, each corresponding to one of M computing boxes, to collect logistics device business data, obtaining M logistics device business log sequences and M logistics device business video sequences, where M is a positive integer. It then iterates through these M log sequences and video sequences to extract anomaly feature anchor vectors, obtaining M log anomaly feature anchor vector sequences and M video anomaly feature anchor vector sequences. Based on the M video anomaly feature anchor vector sequences, it performs attention interaction fusion on the M log anomaly feature anchor vector sequences to obtain M log anomaly feature interaction anchor vector sequences. Logistics task work orders are generated based on these M log anomaly feature interaction anchor vector sequences, resulting in a logistics task work order set. Finally, it acquires logistics resource distribution information for the target area and performs intelligent allocation based on the logistics task work order set to obtain a target intelligent allocation scheme.
[0006] A second aspect of this application provides an intelligent logistics task allocation system oriented towards multi-source needs, the system comprising:
[0007] Data Acquisition Module: Interacts with M computing boxes corresponding to M logistics devices distributed in the target area to collect logistics device business data, obtaining M logistics device business work log sequences and M logistics device business video sequences, where M is a positive integer; Anchor Vector Extraction Module: Traverses the M logistics device business work log sequences and the M logistics device business video sequences to extract abnormal feature anchor vectors, obtaining M log abnormal feature anchor vector sequences and M video abnormal feature anchor vector sequences; Fusion Module: Performs attention interaction fusion on the M log abnormal feature anchor vector sequences based on the M video abnormal feature anchor vector sequences, obtaining M log abnormal feature interaction anchor vector sequences; Work Order Generation Module: Generates logistics task work orders based on the M log abnormal feature interaction anchor vector sequences, obtaining a logistics task work order set; Allocation Module: Obtains logistics resource distribution information in the target area, combines it with the logistics task work order set, and performs intelligent allocation to obtain a target intelligent allocation scheme.
[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0009] First, M logistics devices distributed across the target area, each corresponding to one of the M computing boxes, interact to collect logistics device business data, obtaining M logistics device business work log sequences and M logistics device business video sequences, where M is a positive integer. Next, anomaly feature anchor vectors are extracted from the M logistics device business work log sequences and M logistics device business video sequences, obtaining M log anomaly feature anchor vector sequences and M video anomaly feature anchor vector sequences. Further, attention interaction fusion is performed on the M log anomaly feature anchor vector sequences based on the M video anomaly feature anchor vector sequences, obtaining M log anomaly feature interaction anchor vector sequences. Then, logistics task work orders are generated based on the M log anomaly feature interaction anchor vector sequences, obtaining a set of logistics task work orders. Finally, logistics resource distribution information in the target area is obtained, and intelligent allocation is performed in conjunction with the set of logistics task work orders to obtain a target intelligent allocation scheme. This solves the technical problems of uneven logistics task allocation and low resource utilization in existing technologies, achieving efficient matching of tasks and resources and improving the utilization rate of logistics resources. Attached Figure Description
[0010] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0011] Figure 1A schematic flowchart of a method for intelligent allocation of logistics tasks oriented towards multi-source needs, provided for an embodiment of this application;
[0012] Figure 2 This is a schematic diagram of the structure of an intelligent logistics task allocation system for multi-source needs, provided in an embodiment of this application.
[0013] Figure labeling: Data acquisition module 11, anchor vector extraction module 12, fusion module 13, work order generation module 14, allocation module 15. Detailed Implementation
[0014] This application provides a method and system for intelligent allocation of logistics tasks to meet diverse needs, which solves the technical problems of uneven allocation of logistics tasks and low resource utilization in the prior art.
[0015] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.
[0016] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.
[0017] Example 1, as Figure 1 As shown, this application provides an intelligent allocation method for logistics tasks oriented towards multi-source needs, wherein the method includes:
[0018] The system interacts with M logistics devices distributed in the target area, each corresponding to one of the M computing boxes, to collect logistics device business data, obtaining M logistics device business work log sequences and M logistics device business work video sequences, where M is a positive integer.
[0019] In this embodiment, computing boxes are set up for various types of logistics equipment (such as elevators, air conditioners, water pumps, fire-fighting facilities, inspection robots, etc.) distributed in a target area (such as a smart park, hospital, campus, etc.), with each corresponding logistics equipment to achieve distributed deployment and local data collection and processing. The computing box is a micro computing terminal with edge computing capabilities, which can perform high-speed, low-latency data interaction with the corresponding logistics equipment to collect multimodal data of the logistics equipment in real time during business operations.
[0020] Specifically, the computing power box connects to corresponding logistics equipment via communication interfaces (such as RS-485, Ethernet, Wi-Fi, 5G, etc.) and periodically or on demand collects structured business log data (including status records, fault codes, operation events, alarm information, etc.) generated by the equipment, as well as unstructured business video data generated by connected surveillance cameras. During the collection process, the computing power box performs preliminary processing on the raw data through local caching, compression, and preprocessing methods (such as log timestamp synchronization, video keyframe extraction, etc.) before uploading it to the central server or edge collaboration nodes.
[0021] Based on the correspondence between M logistics devices and M computing boxes, M logistics device business log sequences and M logistics device business video sequences can be obtained respectively, where M is a positive integer representing the total number of logistics devices participating in data collection in the target area.
[0022] The abnormal feature anchor vectors are extracted by traversing the M logistics equipment business work log sequences and the M logistics equipment business work video sequences to obtain M log abnormal feature anchor vector sequences and M video abnormal feature anchor vector sequences.
[0023] By traversing and processing the collected M logistics equipment business log sequences and M logistics equipment business video sequences, abnormal feature information related to logistics tasks is extracted and anchored in the form of structured vectors to generate M log abnormal feature anchor vector sequences and M video abnormal feature anchor vector sequences.
[0024] For each log sequence of logistics equipment, a pre-defined text anomaly feature recognizer is used for processing. This text anomaly feature recognizer, based on a natural language processing model (such as BERT, BiLSTM-CRF, Transformer, etc.), combined with a domain knowledge dictionary and anomaly template library, performs semantic parsing and feature labeling on key fields in the logs (such as equipment status codes, operation event types, fault identifiers, anomaly keywords, etc.), extracts feature anchors reflecting abnormal states, and converts them into corresponding log anomaly feature anchor vectors. By traversing M log sequences, M log anomaly feature anchor vector sequences are ultimately obtained.
[0025] Similarly, for each video sequence of logistics equipment, a pre-defined video anomaly feature recognizer is used for processing. The video anomaly feature recognizer can employ a deep learning-based visual model (such as YOLO, 3D CNN, ViT, I3D, etc.) to extract keyframes or key segments corresponding to abnormal events (such as equipment shaking, liquid leakage, abnormal personnel operation, flames and smoke, etc.) from video clips, and further convert them into video anomaly feature anchor vectors with spatial-temporal feature representation capabilities. By traversing M video sequences, a final sequence of M video anomaly feature anchor vectors is obtained.
[0026] Furthermore, by traversing the M logistics equipment business log sequences and the M logistics equipment business video sequences, abnormal feature anchor vectors are extracted to obtain M log abnormal feature anchor vector sequences and M video abnormal feature anchor vector sequences, including:
[0027] A text anomaly feature recognizer and a video anomaly feature recognizer are pre-built; the text anomaly feature recognizer is used to extract anomaly feature anchor vectors from the M logistics equipment business work log sequences to obtain the M log anomaly feature anchor vector sequences; the video anomaly feature recognizer is used to extract anomaly feature anchor vectors from the M logistics equipment business work video sequences to obtain the M video anomaly feature anchor vector sequences.
[0028] Preferably, a pre-built text anomaly feature recognizer and a video anomaly feature recognizer are included. The text anomaly feature recognizer, built using natural language processing techniques, identifies anomalous events in log data. Specifically, it includes a word embedding layer (e.g., Word2Vec, BERT Embedding), a sequence modeling layer (e.g., Bi-LSTM, Transformer Encoder), and a classification layer. It can identify anomaly patterns such as fault codes, warning states, operational anomalies, and timeout events in equipment logs based on predefined anomaly type labels or weakly supervised rules, and map anomaly fragments into structured vector representations to obtain an anomaly feature anchor vector sequence corresponding to each log sequence. The video anomaly feature recognizer, built using computer vision and temporal modeling methods, identifies video anomaly information during the operation of logistics equipment. Specifically, it includes a video frame extraction module and an image / video object detection model (e.g., YOLOv5, Faster). Video anomaly feature recognizers can locate key abnormal scenes in equipment operation, such as smoke, liquid spills, violent equipment shaking, and unauthorized personnel approaching, and map the detected abnormal key frames or video segments into high-dimensional feature vectors to form a video anomaly feature anchor vector sequence.
[0029] The text anomaly feature recognizer is used to parse and extract features from each of the M logistics equipment business log sequences, resulting in M log anomaly feature anchor vector sequences. The video anomaly feature recognizer is used to process and detect anomalies in each of the M logistics equipment business video sequences, resulting in M video anomaly feature anchor vector sequences.
[0030] Attention interaction fusion is performed on the M video anomaly feature anchor vector sequences to obtain M log anomaly feature interaction anchor vector sequences.
[0031] In this embodiment, an attention-based interactive fusion mechanism is used to process M video anomaly feature anchor vector sequences and M log anomaly feature anchor vector sequences, ultimately obtaining M log anomaly feature interactive anchor vector sequences. Specifically, firstly, the video anomaly feature anchor vector sequences and log anomaly feature anchor vector sequences corresponding to each logistics device are time-synchronized and matched. A timestamp alignment strategy is used to establish associations between anchor pairs within similar time periods. Then, an interaction computation model based on the attention mechanism is constructed, using log anchor vectors as query terms and video anchor vectors as keys and values. Attention coefficients are calculated to obtain the attention weight of each log anchor in the video feature space. Further, the corresponding video anchor vectors are weighted and fused using this weight coefficient. The result is then fused with the original log anchor vectors to obtain a fused vector containing multimodal anomaly feature semantics. Finally, for each logistics device, a fused log anomaly feature interactive anchor vector sequence is generated, providing a precise multi-source anomaly representation foundation for subsequent anomaly correlation analysis and logistics task work order generation.
[0032] Furthermore, based on the M video anomaly feature anchor vector sequences, attention interaction fusion is performed on the M log anomaly feature anchor vector sequences to obtain M log anomaly feature interaction anchor vector sequences, including:
[0033] Simultaneous anchor vector mapping interaction coefficient identification is performed on the M video anomaly feature anchor vector sequences and the M log anomaly feature anchor vector sequences to obtain M mapping interaction coefficient sequences; M attention interaction fusion matrices are constructed based on the M mapping interaction coefficient sequences; the M attention interaction fusion matrices are used to perform interaction fusion on the M log anomaly feature interaction anchor vector sequences to obtain M log anomaly feature interaction anchor vector sequences.
[0034] Preferably, the similarity between the anchor vectors of each corresponding log anomaly feature anchor vector sequence and the video anomaly feature anchor vector sequence is first calculated. Similarity measures such as cosine similarity, Euclidean distance, or dot product-based similarity are used to evaluate the similarity of anchor pairs within the same time window, resulting in M mapping interaction coefficient sequences. Next, based on each mapping interaction coefficient sequence, M corresponding attention interaction fusion matrices are constructed. Specifically, each set of interaction coefficient sequences is normalized to ensure all coefficient values are within a uniform dimension, guaranteeing reasonable weight allocation. The normalized interaction coefficients are then filled into the corresponding upper triangular matrices according to their chronological order in the original sequence, forming a temporally ordered fusion structure that preserves the feature evolution trend. Subsequently, the attention interaction fusion matrices are used to perform weighted fusion processing on the log anomaly feature anchor vector sequences. This involves multiplying each log anchor vector by its corresponding weight factor in the video interaction matrix, summing the results, and combining this with the original log anchor vectors. A new fusion vector is then generated through feature concatenation or residual fusion, ultimately obtaining a log anomaly feature interaction anchor vector sequence containing video interaction semantics.
[0035] Based on the sequence of interactive anchor vectors of M log anomaly features, logistics task work orders are generated to obtain a set of logistics task work orders.
[0036] Furthermore, logistics task work orders are generated based on the M log anomaly feature interaction anchor point vector sequences, resulting in a set of logistics task work orders, including:
[0037] The M log anomaly feature interaction anchor vector sequences are traversed to perform anomaly correlation iterative analysis to obtain M target anomaly iterative memories; based on the M target anomaly iterative memories, logistics task work orders are identified to obtain a set of logistics task work orders.
[0038] Specifically, the process involves traversing M log anomaly feature interaction anchor vector sequences, combining temporal and spatial correlation information between vectors, and the fused multimodal feature representation to perform anomaly correlation iterative analysis. This can be based on recurrent neural networks (RNNs) or graph attention networks (GATs). For each log vector sequence, an anomaly evolution path is constructed, and through multiple iterations, feature fragments highly correlated with the target anomaly event are gradually extracted from the original anchor vectors and stored in a dynamically updated anomaly memory module, thus forming M target anomaly iterative memories. These memory sequences can effectively characterize the potential problem types, occurrence trends, and historical co-occurrence relationships of current logistics equipment, providing feature support for anomaly determination.
[0039] Based on the obtained M target anomaly iterative memories, combined with historical work order tag information or preset work order classification rules, the task identification model (such as a task matcher based on Transformer or BERT) is used to determine the task type of the anomaly features, identify the corresponding logistics task work order content, such as equipment failure, environmental anomaly, safety hazard, material shortage, etc., and output structured work order information, including task name, task type, processing priority, associated equipment ID, anomaly description summary, timestamp and geographical location information, etc., to form a logistics task work order.
[0040] Furthermore, by traversing the M log anomaly feature interaction anchor vector sequences and performing anomaly correlation iterative analysis, M target anomaly iterative memories are obtained, including:
[0041] Each of the M log anomaly feature interaction anchor vector sequences is used to perform anomaly correlation iterative extraction on the M second log anomaly feature interaction anchor vectors to obtain M first anomaly iterative memories. Based on the M first anomaly iterative memories, anomaly correlation iterative extraction is performed on the M log anomaly feature interaction anchor vector sequences to obtain M second anomaly iterative memories. Similarly, based on the M second anomaly iterative memories, anomaly correlation iterative analysis is performed on the M log anomaly feature interaction anchor vector sequences to obtain the M target anomaly iterative memories.
[0042] First, from each of the M log anomaly feature interaction anchor vector sequences, the first log anomaly feature interaction anchor vector for the current stage is extracted and used as the initial reference vector set. For each reference vector, an anomaly association strategy based on similarity calculation (e.g., based on cosine similarity, Euclidean distance, or a custom weighted association function) is used to match each subsequent second log anomaly feature interaction anchor vector in the sequence, identifying anomaly feature vectors that are related to the reference vector in terms of content structure or evolution trajectory. The matching results are then combined to construct M first anomaly iterative memories. Next, based on the M first anomaly iterative memories, the corresponding log anomaly feature interaction anchor vector sequences are traversed again, and the vectors that have not yet participated in the previous round of memory construction are judged for their correlation, thereby further extracting anomaly anchor vectors that have potential extension relationships or evolutionary characteristics with the current memory content, forming M second anomaly iterative memories. Following this pattern, in each subsequent iteration, the abnormal memory results generated in the previous iteration are used as input to continue performing deep correlation mining and completion of feature vectors. This gradually expands the coverage of abnormal memories and strengthens the clustering effect of abnormal events until no new highly correlated abnormal anchor points are found in the current iteration, or the preset iteration threshold is reached. Ultimately, M target abnormal iterative memories with high semantic relevance and contextual coherence are obtained, providing high-quality abnormal knowledge support for subsequent logistics task work order identification and intelligent allocation.
[0043] Furthermore, this includes:
[0044] The correlation degree of anchor vector elements is determined by using M first log anomaly feature interaction anchor vectors to M second log anomaly feature interaction anchor vectors. When the correlation degree determination result of anchor vector elements meets the preset correlation degree threshold, the anchor vector elements are combined to obtain a set of M first correlation iteration elements. The set of M first correlation iteration elements and the M first log anomaly feature interaction anchor vectors to M second log anomaly feature interaction anchor vectors are respectively added to an initially empty vector to obtain M first anomaly iteration memories.
[0045] Preferably, the correlation degree of anchor vector elements is determined by using M first log anomaly feature interaction anchor vectors and M second log anomaly feature interaction anchor vectors respectively. That is, a dimension-wise similarity analysis is performed on the corresponding vector elements in each pair of first and second log anomaly feature interaction anchor vectors, and the element-level correlation degree is calculated by a set similarity function (such as weighted cosine similarity, Manhattan distance, or a custom vector element mapping function). When the correlation degree determination result meets the preset correlation degree threshold condition (such as the correlation degree value is greater than a certain threshold θ), it is determined that the pair of anchor vectors has a valid semantic or behavioral correlation relationship. Then, the anchor vector elements that meet the condition are combined to generate a set of M first correlation iteration elements, which are used to represent the initial anomaly feature structure that needs to be focused on in subsequent iterations. Subsequently, the above set of M first correlation iteration elements, together with the corresponding M first log anomaly feature interaction anchor vectors and M second log anomaly feature interaction anchor vectors, are added to an initially empty vector container to construct the first round of anomaly pattern memory set that can be used for iterative tracking, namely the M first anomaly iteration memories.
[0046] Obtain the distribution information of logistics resources in the target area, and intelligently allocate them in combination with the set of logistics task work orders to obtain a target intelligent allocation scheme.
[0047] By connecting with logistics resource sensing terminals (such as personnel positioning terminals, equipment status monitoring nodes, material storage and allocation systems, etc.) deployed in the target area, real-time distribution information of logistics resources in the area is collected and summarized to form a resource distribution data structure covering various resource types (including manpower, tools, equipment, materials, etc.); at the same time, the task type, service content, estimated time, urgency and location information of each work order are extracted from the generated logistics task work order set to form a logistics task work order location information set. Based on this, by combining the location information set of logistics task work orders with the current resource distribution data, a multi-objective adaptive scheduling model is constructed. This model considers factors such as minimizing task response time, maximizing resource load balancing, and optimizing task priority matching. All work order tasks to be executed are then bound to resources and adapted to their locations to obtain an initial intelligent allocation scheme. Subsequently, resource utilization analysis is performed on the initial allocation results to determine the rationality of resource configuration and overall efficiency under the current scheme. If the analysis results do not meet the preset performance indicators (e.g., the utilization rate of a certain type of resource is lower than the threshold or the task waiting time is too long), a policy-driven intelligent adjustment mechanism is further activated to optimize and iterate the allocation results. Finally, a target intelligent allocation scheme that meets the requirements of scheduling efficiency and resource utilization balance is generated.
[0048] Furthermore, acquiring the distribution information of logistics resources in the target area, and combining it with the set of logistics task work orders for intelligent allocation, yields a target intelligent allocation scheme, including:
[0049] Extract the work order location information from the logistics task work order set to obtain a logistics task work order location information set; combine the logistics resource distribution information and the logistics task work order location information set to adaptively allocate the logistics task work order set to obtain an initial intelligent allocation scheme; perform resource utilization analysis on the initial intelligent allocation scheme, and if the analysis results do not meet the requirements, adjust the initial intelligent allocation scheme to obtain a target intelligent allocation scheme.
[0050] First, the real-time distribution status of various logistics resources within the target area is collected through the logistics resource management system. These resources include, but are not limited to, the location and availability of logistics personnel, the storage location and usage status of various equipment and tools, and the inventory and distribution of materials, forming a logistics resource distribution information database. Then, information such as the specific execution location, task type, and urgency of each task is extracted from the logistics task work order set, constructing a logistics task work order location information set. Based on these two key pieces of information, an adaptive allocation mechanism is used to perform preliminary resource matching on the logistics task work order set. Taking into account multiple factors such as task urgency, resource allocation cost, path distance, and execution capability matching, an initial intelligent allocation scheme is generated through a multi-objective optimization algorithm. Specifically, the real-time location, availability, and capability parameters of various logistics resources within the target area are matched with the task location and demand information in the logistics task work order location information set to construct a resource-task association matrix. Based on this association matrix, an adaptive scheduling algorithm is adopted to dynamically adjust the matching relationship between tasks and resources, taking into account multiple factors such as task urgency, resource availability, task execution distance, and resource load balancing. Through iterative optimization, the allocation scheme is continuously adjusted to reduce task response time and resource idle rate. Finally, an initial intelligent allocation scheme that can effectively cover all task requirements and has a high resource utilization rate is generated.
[0051] The system performs resource utilization analysis on the initial intelligent allocation scheme, including calculating the workload balance of logistics personnel and equipment, estimating task completion time, and overall resource utilization efficiency. If the analysis results indicate issues such as resource idleness or task response delays, a rule-based and machine learning-based adjustment mechanism is activated to dynamically adjust the resource allocation ratio, task order, or task-resource matching relationship in the allocation scheme, continuously optimizing the allocation effect. Ultimately, the system outputs a target intelligent allocation scheme that meets performance indicators, maximizes resource utilization, and satisfies logistics task requirements, providing scientific and efficient decision support for logistics management.
[0052] Furthermore, this includes:
[0053] According to the preset adjustment strategy, the initial intelligent allocation scheme is randomly adjusted to obtain an adjusted intelligent allocation scheme set; the resource utilization rate of the adjusted intelligent allocation scheme set is analyzed by traversing the adjusted intelligent allocation scheme set to obtain an adjusted intelligent allocation scheme resource utilization rate set; it is determined whether there is an adjusted intelligent allocation scheme resource utilization rate in the adjusted intelligent allocation scheme resource utilization rate set that is greater than or equal to the resource utilization rate of the initial intelligent allocation scheme. If so, the adjusted intelligent allocation scheme corresponding to the maximum value in the adjusted intelligent allocation scheme resource utilization rate set is taken as the target intelligent allocation scheme.
[0054] Specifically, according to a preset adjustment strategy, the number of task orders assigned to each processor in the initial intelligent allocation scheme is appropriately increased or decreased. Simultaneously, the matching order and specific allocation relationship between tasks and resources may be adjusted, generating multiple different adjustment schemes. These adjustment scheme sets are automatically and randomly generated by the system to ensure that various possible resource allocation methods are explored within a reasonable range. Subsequently, the system performs resource utilization analysis on each scheme in the adjustment scheme set, mainly by calculating indicators such as the workload balance of logistics personnel and equipment, task response time, resource idle rate, and overall task completion efficiency, forming corresponding resource utilization values and constructing a resource utilization set of adjusted intelligent allocation schemes. The system further compares the resource utilization indicators of all adjustment schemes in this set to determine whether there is an adjustment scheme whose resource utilization is greater than or equal to that of the initial intelligent allocation scheme. If so, the adjustment scheme with the highest resource utilization is selected as the final target intelligent allocation scheme to achieve optimal allocation and maximum utilization of logistics resources, ensuring efficient completion of logistics tasks and reasonable resource scheduling.
[0055] Furthermore, the preset adjustment strategy is to increase or decrease the number of work orders processed by each processor in the initial intelligent allocation scheme.
[0056] The preset adjustment strategy involves appropriately increasing or decreasing the number of work orders assigned to each processor in the initial intelligent allocation scheme. In other words, the system randomly increases or decreases the number of work orders handled by a particular processor based on certain adjustment ranges and rules, thereby fine-tuning the entire task allocation scheme.
[0057] In summary, the embodiments of this application have at least the following technical effects:
[0058] First, M logistics devices distributed across the target area, each corresponding to one of the M computing boxes, interact to collect logistics device business data, obtaining M logistics device business work log sequences and M logistics device business video sequences, where M is a positive integer. Next, anomaly feature anchor vectors are extracted from the M logistics device business work log sequences and M logistics device business video sequences, obtaining M log anomaly feature anchor vector sequences and M video anomaly feature anchor vector sequences. Further, attention interaction fusion is performed on the M log anomaly feature anchor vector sequences based on the M video anomaly feature anchor vector sequences, obtaining M log anomaly feature interaction anchor vector sequences. Then, logistics task work orders are generated based on the M log anomaly feature interaction anchor vector sequences, obtaining a set of logistics task work orders. Finally, logistics resource distribution information in the target area is obtained, and intelligent allocation is performed in conjunction with the set of logistics task work orders to obtain a target intelligent allocation scheme. This solves the technical problems of uneven logistics task allocation and low resource utilization in existing technologies, achieving efficient matching of tasks and resources and improving the utilization rate of logistics resources.
[0059] Example 2, based on the same inventive concept as the intelligent allocation method for logistics tasks oriented towards multi-source needs in the foregoing examples, such as... Figure 2 As shown, this application provides an intelligent logistics task allocation system for multi-source needs, wherein the system includes:
[0060] Data Acquisition Module 11: Interacts with M computing boxes corresponding to M logistics equipment distributed in the target area to collect logistics equipment business data, obtaining M logistics equipment business work log sequences and M logistics equipment business video sequences, where M is a positive integer; Anchor Vector Extraction Module 12: Traverses the M logistics equipment business work log sequences and the M logistics equipment business video sequences to extract abnormal feature anchor vectors, obtaining M log abnormal feature anchor vector sequences and M video abnormal feature anchor vector sequences; Fusion Module 13: Performs attention interaction fusion on the M log abnormal feature anchor vector sequences based on the M video abnormal feature anchor vector sequences, obtaining M log abnormal feature interaction anchor vector sequences; Work Order Generation Module 14: Generates logistics task work orders based on the M log abnormal feature interaction anchor vector sequences, obtaining a logistics task work order set; Allocation Module 15: Obtains logistics resource distribution information in the target area, combines it with the logistics task work order set for intelligent allocation, and obtains a target intelligent allocation scheme.
[0061] Furthermore, the anchor vector extraction module 12 is used to perform the following method:
[0062] A text anomaly feature recognizer and a video anomaly feature recognizer are pre-built; the text anomaly feature recognizer is used to extract anomaly feature anchor vectors from the M logistics equipment business work log sequences to obtain the M log anomaly feature anchor vector sequences; the video anomaly feature recognizer is used to extract anomaly feature anchor vectors from the M logistics equipment business work video sequences to obtain the M video anomaly feature anchor vector sequences.
[0063] Furthermore, the fusion module 13 is used to perform the following methods:
[0064] Simultaneous anchor vector mapping interaction coefficient identification is performed on the M video anomaly feature anchor vector sequences and the M log anomaly feature anchor vector sequences to obtain M mapping interaction coefficient sequences; M attention interaction fusion matrices are constructed based on the M mapping interaction coefficient sequences; the M attention interaction fusion matrices are used to perform interaction fusion on the M log anomaly feature interaction anchor vector sequences to obtain M log anomaly feature interaction anchor vector sequences.
[0065] Furthermore, the work order generation module 14 is used to execute the following method:
[0066] The M log anomaly feature interaction anchor vector sequences are traversed to perform anomaly correlation iterative analysis to obtain M target anomaly iterative memories; based on the M target anomaly iterative memories, logistics task work orders are identified to obtain a set of logistics task work orders.
[0067] Furthermore, the work order generation module 14 is used to execute the following method:
[0068] Each of the M log anomaly feature interaction anchor vector sequences is used to perform anomaly correlation iterative extraction on the M second log anomaly feature interaction anchor vectors to obtain M first anomaly iterative memories. Based on the M first anomaly iterative memories, anomaly correlation iterative extraction is performed on the M log anomaly feature interaction anchor vector sequences to obtain M second anomaly iterative memories. Similarly, based on the M second anomaly iterative memories, anomaly correlation iterative analysis is performed on the M log anomaly feature interaction anchor vector sequences to obtain the M target anomaly iterative memories.
[0069] Furthermore, the work order generation module 14 is used to execute the following method:
[0070] The correlation degree of anchor vector elements is determined by using M first log anomaly feature interaction anchor vectors to M second log anomaly feature interaction anchor vectors. When the correlation degree determination result of anchor vector elements meets the preset correlation degree threshold, the anchor vector elements are combined to obtain a set of M first correlation iteration elements. The set of M first correlation iteration elements and the M first log anomaly feature interaction anchor vectors to M second log anomaly feature interaction anchor vectors are respectively added to an initially empty vector to obtain M first anomaly iteration memories.
[0071] Furthermore, the allocation module 15 is used to perform the following method:
[0072] Extract the work order location information from the logistics task work order set to obtain a logistics task work order location information set; combine the logistics resource distribution information and the logistics task work order location information set to adaptively allocate the logistics task work order set to obtain an initial intelligent allocation scheme; perform resource utilization analysis on the initial intelligent allocation scheme, and if the analysis results do not meet the requirements, adjust the initial intelligent allocation scheme to obtain a target intelligent allocation scheme.
[0073] Furthermore, the allocation module 15 is used to perform the following method:
[0074] According to the preset adjustment strategy, the initial intelligent allocation scheme is randomly adjusted to obtain an adjusted intelligent allocation scheme set; the resource utilization rate of the adjusted intelligent allocation scheme set is analyzed by traversing the adjusted intelligent allocation scheme set to obtain an adjusted intelligent allocation scheme resource utilization rate set; it is determined whether there is an adjusted intelligent allocation scheme resource utilization rate in the adjusted intelligent allocation scheme resource utilization rate set that is greater than or equal to the resource utilization rate of the initial intelligent allocation scheme. If so, the adjusted intelligent allocation scheme corresponding to the maximum value in the adjusted intelligent allocation scheme resource utilization rate set is taken as the target intelligent allocation scheme.
[0075] Furthermore, the allocation module 15 is used to perform the following method:
[0076] The preset adjustment strategy is to increase or decrease the number of work orders processed by each processor in the initial intelligent allocation scheme.
[0077] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.
[0078] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.
[0079] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.
Claims
1. A method for intelligent allocation of logistics tasks oriented towards multi-source needs, characterized in that, The method includes: The system interacts with M logistics devices distributed in the target area, each corresponding to one of the M computing power boxes, to collect logistics device business data, and obtains M logistics device business work log sequences and M logistics device business video sequences, where M is a positive integer; The abnormal feature anchor vectors are extracted by traversing the M logistics equipment business work log sequences and the M logistics equipment business work video sequences to obtain M log abnormal feature anchor vector sequences and M video abnormal feature anchor vector sequences. Attention interaction fusion is performed on the M video anomaly feature anchor vector sequences to obtain M log anomaly feature interaction anchor vector sequences. Based on the sequence of interactive anchor vectors of M log anomaly features, a logistics task work order is generated to obtain a set of logistics task work orders. Obtain the distribution information of logistics resources in the target area, and intelligently allocate them in combination with the set of logistics task work orders to obtain the target intelligent allocation scheme; Specifically, logistics task work orders are generated based on M log anomaly feature interaction anchor point vector sequences, resulting in a set of logistics task work orders, including: By traversing the M log anomaly feature interaction anchor vector sequences, anomaly correlation iterative analysis is performed to obtain M target anomaly iterative memories; Based on the M target anomaly iterative memories, the logistics task work order is identified to obtain a set of logistics task work orders; Specifically, anomaly correlation iterative analysis is performed by traversing the M log anomaly feature interaction anchor point vector sequences to obtain M target anomaly iterative memories, including: The M log anomaly feature interaction anchor vectors are used to perform anomaly correlation iterative extraction on the M second log anomaly feature interaction anchor vectors in the M log anomaly feature interaction anchor vector sequence to obtain M first anomaly iterative memories, and the first log anomaly feature interaction anchor vectors are used as the initial reference vector set. Based on the M first abnormal iterative memories, abnormal correlation iterative extraction is performed on the M log abnormal feature interaction anchor vector sequences to obtain M second abnormal iterative memories; Similarly, based on the M second anomaly iterative memories, anomaly correlation iterative analysis is performed on the M log anomaly feature interaction anchor vector sequences to obtain the M target anomaly iterative memories; The correlation degree of anchor vector elements is determined by M first log anomaly feature interaction anchor vectors and M second log anomaly feature interaction anchor vectors respectively. When the correlation degree determination result of anchor vector elements meets the preset correlation degree threshold, the anchor vector elements are combined to obtain a set of M first correlation iteration elements. The correlation degree determination of anchor vector elements is to perform a dimension-wise similarity analysis on the corresponding vector elements in each pair of first and second log anomaly feature interaction anchor vectors and calculate their element-level correlation degree. The M sets of first associated iterative elements and the M sets of first log anomaly feature interaction anchor vectors and the M sets of second log anomaly feature interaction anchor vectors are respectively added to the initially empty vector to obtain the M first anomaly iterative memories. This includes obtaining the distribution information of logistics resources in the target area, combining it with the set of logistics task work orders for intelligent allocation, and obtaining a target intelligent allocation scheme, including: Extract the work order location information of the logistics task work order set to obtain the logistics task work order location information set; By combining the logistics resource distribution information and the logistics task work order location information set, the logistics task work order set is adaptively allocated to obtain an initial intelligent allocation scheme. The initial intelligent allocation scheme is analyzed for resource utilization. If the analysis results do not meet the requirements, the initial intelligent allocation scheme is adjusted to obtain the target intelligent allocation scheme. According to the preset adjustment strategy, the initial intelligent allocation scheme is randomly adjusted to obtain an adjusted intelligent allocation scheme set; By traversing the set of intelligent allocation schemes for adjustment, resource utilization analysis is performed to obtain a set of resource utilization rates for intelligent allocation schemes for adjustment. Determine whether there exists an adjusted intelligent allocation scheme resource utilization rate in the set of adjusted intelligent allocation scheme resource utilization rates that is greater than or equal to the resource utilization rate of the initial intelligent allocation scheme. If so, the adjusted intelligent allocation scheme corresponding to the maximum value in the set of adjusted intelligent allocation scheme resource utilization rates is taken as the target intelligent allocation scheme. The preset adjustment strategy is to increase or decrease the number of work orders processed by each processor in the initial intelligent allocation scheme.
2. The intelligent allocation method for logistics tasks oriented towards multi-source needs as described in claim 1, characterized in that, The M logistics equipment business log sequences and M logistics equipment business video sequences are traversed to extract anomaly feature anchor vectors, resulting in M log anomaly feature anchor vector sequences and M video anomaly feature anchor vector sequences, including: Pre-built text anomaly feature recognizers and video anomaly feature recognizers; Anomaly feature anchor vectors are extracted from the M logistics equipment business log sequences using a text anomaly feature recognizer to obtain the M log anomaly feature anchor vector sequences. Anomaly feature anchor vectors are extracted from the M logistics equipment operation video sequences using a video anomaly feature identifier, resulting in M video anomaly feature anchor vector sequences.
3. The intelligent allocation method for logistics tasks oriented towards multi-source needs as described in claim 1, characterized in that, Based on the M video anomaly feature anchor vector sequences, attention interaction fusion is performed on the M log anomaly feature anchor vector sequences to obtain M log anomaly feature interaction anchor vector sequences, including: Simultaneous anchor vector mapping interaction coefficient identification is performed on the M video anomaly feature anchor vector sequences and the M log anomaly feature anchor vector sequences to obtain M mapping interaction coefficient sequences. M attention interaction fusion matrices are constructed based on the M mapping interaction coefficient sequences; The M attention interaction fusion matrices are used to perform interaction fusion on the M log anomaly feature interaction anchor vector sequences to obtain M log anomaly feature interaction anchor vector sequences.
4. A logistics task intelligent allocation system oriented towards multi-source needs, characterized in that, The system is used to implement the intelligent allocation method for logistics tasks oriented towards multi-source needs as described in any one of claims 1-3, the system comprising: Data acquisition module: Interacts with M computing boxes corresponding to M logistics devices distributed in the target area to collect logistics device business data, and obtains M logistics device business work log sequences and M logistics device business video sequences, where M is a positive integer; Anchor vector extraction module: Traverse the M logistics equipment business work log sequences and the M logistics equipment business work video sequences to extract abnormal feature anchor vectors, and obtain M log abnormal feature anchor vector sequences and M video abnormal feature anchor vector sequences; Fusion module: Based on the M video anomaly feature anchor vector sequences, the M log anomaly feature anchor vector sequences are fused with attention interaction to obtain M log anomaly feature interaction anchor vector sequences; Work order generation module: Generates logistics task work orders based on M log anomaly feature interaction anchor vector sequences, and obtains a set of logistics task work orders; Allocation module: Obtains the distribution information of logistics resources in the target area, and performs intelligent allocation in combination with the set of logistics task work orders to obtain the target intelligent allocation scheme.