Mobile storage and charging robot remote scheduling and path planning system based on Internet of Things

Abstract

The invention discloses a mobile storage and charging robot remote scheduling and path planning system based on the Internet of Things, and relates to the technical field of robot control. Comprising a path dependency modeling module, a path conflict analysis module, a resource dependency graph construction module, a decoupling rearrangement scheduling module, a time sequence offset evaluation module and a path weight regulation and control module, and obtaining a path node sequence and an access time period of a current to-be-executed task of each mobile storage and charging robot, and generating a task path pre-occupation graph. By constructing the path dependence model, the conflict prediction mechanism and the task decoupling rearrangement strategy, accurate identification and effective intervention of path conflicts and resource deadlocks in the multi-robot scheduling process are realized, the stability of the scheduling system and the task execution continuity are improved, efficient completion of energy supply is ensured, and the scheduling efficiency is improved. And the operation efficiency and safety of the system are obviously optimized.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a remote scheduling and path planning system for a mobile storage and charging robot based on the Internet of Things. Background Art

[0002] The IoT-based remote scheduling and path planning system for mobile energy storage and charging robots refers to an intelligent management platform built on the perception, communication, and control capabilities of the IoT, designed to remotely coordinate and dynamically plan paths for the operating tasks of multiple mobile energy storage and charging integrated robots. By deploying multiple types of edge sensing nodes, the system collects key data such as environmental conditions, obstacle distribution, power levels, and path accessibility in the operating area in real time, and uploads this data to the cloud or edge server through the IoT network for centralized processing and intelligent decision-making. Based on dynamic accessibility analysis, path risk assessment, energy consumption prediction, and task priority sorting, the system generates the optimal path plan for collaborative operations between multiple robots, and realizes remote control of each robot, issuance of scheduling instructions, and real-time path updates, thereby ensuring the efficient operation and safe obstacle avoidance of multiple devices in complex environments, and improving overall operational efficiency and energy management.

[0003] Existing technologies suffer from the following shortcomings: In complex and dynamic environments, existing IoT-based mobile storage and charging robot scheduling systems typically rely on path prediction models to synchronously plan the routes of multiple robots. However, due to uncertainties such as temporary obstacles, frequent path changes, and sensor information latency, existing path prediction algorithms are prone to misjudging dynamic traffic states when generating routes. This can lead to multiple robots being incorrectly scheduled to interdependent path nodes, creating a logical path resource loop. Specifically, each robot must wait for other robots to release critical path nodes before continuing, but these other robots are also in a similar waiting state, causing all robots to enter a mutually blocked "waiting loop." Under this structure, path resources are never released, preventing robots from completing path transfers, ultimately leading to system-level scheduling deadlock. Once this problem occurs, it can severely impact overall mission execution efficiency, leading to failures in critical energy resupply, and even causing equipment congestion, communication resource exhaustion, and overall scheduling system failure, posing significant safety and operational risks.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0005] The purpose of the present invention is to provide a remote scheduling and path planning system for mobile storage and charging robots based on the Internet of Things. By constructing a path dependency model, a conflict prediction mechanism and a task decoupling and rescheduling strategy, it can achieve accurate identification and effective intervention of path conflicts and resource deadlocks in the multi-robot scheduling process, improve the stability of the scheduling system and the continuity of task execution, ensure efficient energy replenishment, and significantly optimize the system operation efficiency and safety, so as to solve the problems in the above-mentioned background technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a remote scheduling and path planning system for mobile storage and charging robots based on the Internet of Things, comprising a path dependency modeling module, a path conflict analysis module, a resource dependency graph construction module, a decoupling and rescheduling module, a timing offset assessment module, and a path weight control module: The path dependency modeling module builds a task path dependency extraction model, obtains the path node sequence and access time period of each mobile storage and charging robot's current task to be executed, and generates a task path pre-occupancy map; The path conflict analysis module maps the pre-occupancy graphs of each task path onto a unified time axis, identifies the access overlap of path nodes in each time period, constructs a set of path intersection nodes, calculates the path intersection density, and generates a path conflict distribution map. The resource dependency graph construction module builds a resource access dependency graph between tasks based on the path conflict distribution graph and the path intersection node set, identifies whether there is a closed access dependency chain, and if so, extracts the task set and generates a loop identification set; The decoupling and rescheduling module decouples and reschedules tasks based on the loop identification set, adjusts the execution order of some mobile storage and charging robots or modifies the path access time, generates an updated task timing diagram, and verifies whether the loop structure is broken; The timing offset evaluation module compares the updated task timing diagram with the original task time information, calculates the difference between the start and end time of each task, and outputs the task timing offset data; The path weight control module combines the path conflict density and task timing offset results to construct a path control weight matrix, dynamically adjusts the access priority and release time of path nodes, and guides path selection to avoid high-conflict and high-offset nodes.

[0007] Preferably, building a task path dependency extraction model includes the following steps: Obtain the starting position, target position, expected execution time window and task priority of each mobile storage and charging robot's current task; Call the path planning engine to generate a path that includes a sequence of path nodes and the entry and exit times of each node; The path nodes and their time information are constructed as node time label tuples, which include path node identification, entry timestamp, exit timestamp and task priority factor; All node time label tuples are aggregated into a task path preemption set and uniformly constructed into a task path preemption graph.

[0008] Preferably, mapping the task path pre-occupancy graph uniformly to a unified time axis and generating a path conflict distribution graph comprises the following steps: Perform temporal reordering on the entry and exit timestamps in each node's time label tuple to complete unified timeline mapping. In each time window, the path nodes visited by multiple tasks are counted and cross-feature verification is performed to determine the path intersection nodes; Construct a set of path intersection nodes and calculate the visit frequency and path intersection density factor of each node; A path conflict distribution graph with time axis and space nodes as coordinates is generated based on the path intersection node set.

[0009] Preferably, constructing a resource access dependency graph and generating a loop identification set includes the following steps: Based on the path conflict distribution graph and the path intersection node set, For task pairs with overlapping access relationships and waiting dependencies, a resource access dependency graph is constructed with tasks as nodes and dependency relationships as edges. Identify graph traversal loops on the dependency graph to detect whether there is a closed access dependency chain; If there is a closed dependency chain, the corresponding task number, intersection node, and conflict time period are extracted to generate a loop identifier; All loop identifications are aggregated into a loop identification set for subsequent scheduling optimization processing.

[0010] Preferably, performing task decoupling and rearrangement based on the loop identification set includes the following steps: Extract the loop task set from the loop identification set, evaluate the impact of each task on the decoupling effect based on the scheduling intervention influence factor model, and determine the key tasks; Adjust the execution order of key tasks or modify the path access time, and introduce a time control weight coefficient to limit the maximum offset range; Write the adjustment results into the task scheduling table, generate an updated task timing diagram, and reconstruct the task path pre-occupancy diagram and resource access dependency diagram; If the loop structure is not completely eliminated, the decoupling process is executed repeatedly until the loop is eliminated, and the decoupling scheme is dynamically optimized using the adaptive decoupling scheduling strategy set.

[0011] Preferably, outputting the timing offset data corresponding to each task includes the following steps: Extract the start and end times of tasks before and after decoupling, and construct the original task time mapping table and the adjusted task time mapping table; Calculate the offset between the start and end time of each task, and generate a task timing offset data tuple containing the offset value, offset direction and rhythm status; Aggregate all timing offset data into a task timing offset matrix and set the offset severity threshold to highlight it; Generate a scheduling rhythm impact assessment report and store the results in the scheduling history database for subsequent strategy optimization.

[0012] Preferably, constructing the pathway control weight matrix includes the following steps: Extract the path node conflict density in the path conflict distribution map and the offset data in the task timing offset matrix; According to the conflict density of each path node and the timing deviation degree of the associated tasks, the access priority level and release time parameters are set; Generate a path control weight matrix with path nodes as indexes, and record the priority and release sequence of each node in different time windows; During the path selection process, the weight matrix is ​​referenced to dynamically guide the robot to avoid high-conflict and high-offset path nodes.

[0013] In the above technical solution, the technical effects and advantages provided by the present invention are: By introducing a multi-level dynamic scheduling mechanism involving path dependency modeling, path conflict prediction, resource dependency graph construction, and task decoupling and rescheduling, this invention enables the early identification, precise location, and effective intervention of path conflicts and scheduling deadlock risks faced by mobile storage and charging robots in complex environments, significantly improving the robustness and execution continuity of system scheduling. Compared to existing technologies, this solution not only possesses high-precision resource conflict identification capabilities for asynchronous multi-robot scheduling, but also adaptively optimizes path priorities and access timing by combining timing offset assessment and path control weights, thereby avoiding system-level deadlocks caused by resource loop dependencies, ensuring smooth progress of robot tasks and timely completion of energy replenishment, and improving overall operational efficiency and system scheduling stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction to the drawings required for use in the embodiments will be given below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0015] Figure 1 This is a module diagram of the remote scheduling and path planning system for mobile storage and charging robots based on the Internet of Things of the present invention. DETAILED DESCRIPTION

[0016] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0017] The present invention provides Figure 1 The IoT-based remote scheduling and path planning system for mobile storage and charging robots shown in the figure includes: The path dependency modeling module builds a task path dependency extraction model. For each task currently to be performed by the mobile storage and charging robot, it extracts the complete path node sequence corresponding to the task and the access time period of each node, and generates a task path pre-occupancy map to identify the path nodes to be occupied and their corresponding access time periods. To effectively predict and intervene in path resource conflicts during the remote scheduling of multiple mobile storage and charging robots, we first need to build a task path dependency extraction model to generate a path resource pre-occupancy information map for each mobile storage and charging robot's current pending tasks. This is achieved through the following steps: For each mobile charging and storage robot, its currently assigned task to be executed is obtained. The task information should include the task starting position, task target position, expected execution time window and task priority. After obtaining the basic task information, the internally integrated path planning engine is called to generate an optimal task path for each robot based on the environmental status data in the current operating area (including obstacle distribution, power station location, access restrictions, etc.) and the current position and motion characteristics of each mobile charging and storage robot (such as maximum speed, acceleration, turning radius, etc.). The path should be presented in the form of a "path node sequence", and each path node represents a spatial coordinate point or access control point that the robot needs to visit in sequence during the execution of the task. To ensure the timing controllability of the scheduling, the expected entry time and exit time of each path node will be calculated at the same time to form a node time period comparison table for the subsequent construction of the path access time window.

[0018] Based on the aforementioned path node sequence and its time period information, a task path resource pre-occupancy structure diagram is constructed. This structure diagram uses "task number" as an index item, decomposing the task path corresponding to each mobile storage and charging robot into multiple "path node-access time period" pairs and mapping them into a two-dimensional space-time coordinate system. To enhance the data adaptability and scalability of the model, a structured parameter "node time label tuple" is introduced. This label tuple includes four fields: path node identifier, entry timestamp, exit timestamp, and task priority factor. The task priority factor is used to identify the priority reference value of the path access operation in global scheduling, facilitating priority-based regulation and rescheduling when subsequent resource conflicts occur. All path node time label tuples constitute the current task path pre-occupancy set, which serves as the basic unit of the task path dependency model.

[0019] The "node time label tuple" is a structured data unit defined for path resource pre-occupancy management and conflict prediction. It accurately describes a mobile storage and charging robot's access behavior to a specific path node while performing a specific task. This label tuple contains four core fields: path node identifier, entry timestamp, exit timestamp, and task priority factor. The path node identifier uniquely identifies the spatial location the robot will pass through. The entry and exit timestamps record the robot's estimated entry and exit times, respectively, and are used to calibrate the node's occupancy period within the scheduling cycle. The task priority factor quantifies the task's scheduling level to facilitate priority determination in the event of conflicts. The core purpose of introducing the "node time label tuple" is to transform path access behavior from a traditional static geometric path representation into a dynamic resource block model with a temporal dimension and scheduling value. This allows for accurate modeling of path resources in both temporal and spatial dimensions, conflict location, and access priority control. This is a key foundation for constructing a task path pre-occupancy graph and subsequent deadlock prediction mechanisms.

[0020] After extracting the path dependency data for a single task, the path pre-occupancy collection for all mobile storage robot tasks is centrally stored and uniformly managed to form a "task path pre-occupancy map" for the scheduling cycle. This map uses time series as the main axis and path nodes as spatial mapping points to record all path nodes that the robots are scheduled to visit in the future and their occupied time periods. To improve the accuracy of subsequent conflict predictions, a dynamic update mechanism is introduced. When the robot task status changes (such as early completion, task cancellation, path replanning, etc.), a local update operation of the path pre-occupancy map is immediately triggered to ensure that the pre-occupancy map always reflects the current actual scheduling expectation status. In addition, the path pre-occupancy map also has an access record status indicator to indicate whether each path node has been actually visited, whether it is in a waiting state, or whether it is a shared node, so as to achieve refined modeling in conflict detection and deadlock prediction.

[0021] After extracting the path dependency data for a single mobile storage and charging robot task, in order to build a system-level scheduling infrastructure, the path reservation sets of all tasks need to be centrally integrated and uniformly managed. This includes the following steps: The node time label tuples generated by each mobile storage and charging robot are classified and processed by task number to ensure that the path node sequence and access time period corresponding to each task are structurally consistent and traceable. The purpose of this step is to provide a unified task-level identification and organizational basis for subsequent data indexing, updating and conflict analysis.

[0022] Normalize the node time label tuples of all tasks to the same time axis, standardize all access times according to a unified time base, and enable the path access data of different tasks to be analyzed in the same time coordinate system; this step is to open up the time dimension for subsequent time sequence overlap detection and intersection analysis, and avoid analysis errors caused by inconsistent time bases of different robots' tasks. All standardized node time label tuples are sorted in time priority order, and the data is organized using a three-layer nested index of "path node → time period → task number" to construct a queryable and updateable graph structure. This step provides an efficient spatiotemporal access query mechanism for subsequent path conflict analysis, while supporting concurrent reading, writing, and real-time updates.

[0023] The above data structure is registered as the "task path pre-occupancy map" within the scheduling cycle, and the map is input into core modules such as path intersection prediction, resource dependency mapping, deadlock detection and task rescheduling as a dynamic resource mapping. The role of this step is to complete the transition from task-level local path abstraction to system-level spatiotemporal resource modeling, providing a complete, dynamic and predictable path resource occupancy view for the multi-robot scheduling system, forming the basic data dependency of the entire intelligent scheduling decision-making process.

[0024] After the path pre-emption graph is constructed, it is passed as an input data source to subsequent analysis units such as the path intersection prediction module and the path loop detection engine. This allows for the spatiotemporal identification of path conflicts and the prediction and assessment of scheduling deadlock risks. Furthermore, the task path dependency extraction model supports integration with edge computing nodes to enable distributed construction of the task path pre-emption graph and local path prediction optimization.

[0025] This step provides the scheduling system with a spatiotemporal basis for predicting path resources. By constructing a task path dependency extraction model, the system extracts the complete path node sequence and access time period for each mobile storage and charging robot task, generating a task path pre-occupancy map that identifies path nodes that will be occupied within a specific timeframe. This map, serving as the core input for scheduling analysis, intuitively reflects the path overlap and resource conflict risks between tasks. It provides high-precision data support for subsequent modules such as path conflict detection, path intersection prediction, resource dependency mapping, and scheduling deadlock prediction. It is a key prerequisite for achieving dynamic multi-robot path coordination and scheduling stability.

[0026] The path conflict analysis module maps the task path pre-occupancy diagrams of each mobile storage and charging robot onto a unified timeline, analyzes the access overlap of path nodes in each time period, constructs a set of path intersection nodes, and counts the access frequency and path intersection density of each path intersection node in the future time period to generate a path conflict distribution map. To effectively identify the risk of path resource competition among multiple mobile storage and charging robots during future mission execution, it is necessary to implement a unified spatiotemporal analysis of path access behavior and conflict pattern identification based on the construction of the task path dependency extraction model. The following steps are used to complete the path intersection prediction and conflict distribution map construction: The task path pre-occupancy graphs corresponding to all mobile storage and charging robots are uniformly mapped to a unified time axis. The specific operation is to re-position the entry timestamp and exit timestamp in each "node time label tuple" based on the global time reference set by the dispatch center, that is, to perform time alignment conversion on the time labels of the original task paths so that the node occupancy periods of all task paths can be represented on the same global time axis. The key to this step is to achieve the "spatiotemporal co-expression" of path access data between different robot tasks, providing a time synchronization basis for subsequent analysis. At the same time, the "time axis alignment window" parameter is introduced to define the minimum granularity and time step of the unified time axis to ensure that the subsequent conflict identification of path access behavior has sufficient resolution and control accuracy.

[0027] The "Timeline Alignment Window" is a customizable time control factor used to precisely align and fine-grainedly analyze the path access data of multiple mobile storage and charging robots on a unified timeline. Its core concept is to divide the global unified timeline into several fixed-length time windows. Each window is a minimum time granularity unit used to discretize the robot's access behavior to path nodes. This parameter provides a standardized time slicing mechanism, allowing path access periods for different tasks to be compared, superimposed, and conflict identified within the same timeframe, enabling precise temporal overlap analysis and access sequence control.

[0028] The specific implementation steps are as follows: When the scheduling system is initialized, a reasonable timeline alignment window length is set, for example, in seconds or milliseconds, and it can be flexibly configured according to the scheduling accuracy requirements of the actual application scenario; Map the entry and exit timestamps of each task path node to the corresponding time window number, that is, convert the continuous time period into a series of discrete window identifiers, thereby forming a path node occupancy table based on time windows; Count the number of tasks accessing the path nodes in each time window, determine whether there are overlapping accesses to the same node by multiple tasks, and identify potential conflicting nodes; Based on factors such as access density and task priority within each time window, the update frequency of the path conflict distribution map and the conflict weight calculation logic are dynamically adjusted to ensure that the conflict identification process has sufficient temporal resolution while adapting to the scheduling requirements of different task rhythms. By introducing the "timeline alignment window" parameter, not only is the robot's task path access behavior uniformly expressed in the temporal dimension, but it also enhances the analysis accuracy and response sensitivity of the conflict identification mechanism in high-concurrency and complex scenarios. This is one of the key parameters that are indispensable for path intersection prediction and task scheduling optimization.

[0029] On the basis of completing the timeline alignment, the access behaviors of all path nodes in each time step are aggregated and statistically analyzed. The specific method is to scan all node time label tuples, extract the path nodes accessed by multiple tasks in the same time window, and identify them as path intersection candidate nodes. On the basis of the existing path conflict analysis technology, in order to improve the accuracy of path intersection node identification and the accuracy of scheduling control, the present invention further introduces a multi-dimensional cross-feature verification mechanism on the basis of identifying the path nodes shared by multiple tasks, and makes a comprehensive judgment on whether the path nodes constitute an actual conflict. Specifically, not only the spatial sharing relationship of the path nodes is identified, but also the following three key dimensions are synchronously verified: First, whether the access directions intersect, that is, whether the entry and exit directions of multiple tasks to the same path node have an intersection of motion trajectories; second, whether the access time overlaps, that is, whether the time windows of multiple tasks when accessing the node partially or completely overlap; third, whether there is a priority conflict in the access order, that is, whether the access plans of multiple tasks at the node cannot be coordinated due to priority sorting. Only when all three of the above conditions are met is the node officially identified as a path intersection node and included in the path intersection node set for subsequent conflict density analysis and scheduling optimization. This mechanism maintains computational efficiency while effectively preventing path nodes that overlap spatially but do not affect each other from being mistakenly identified as conflicting nodes, thereby improving the reliability of path conflict identification and the targeted nature of scheduling interventions.

[0030] All identified path intersection nodes are temporally classified and access density statistics are performed to construct a path intersection node set. This set uses the path node identifier as the basic index, records the access frequency of each intersection node in each future time window, and introduces the "path intersection density factor" as an evaluation indicator to quantify the intensity of each intersection node being accessed simultaneously by multiple tasks within a specific time period. The path intersection density factor comprehensively considers parameters such as the number of access tasks, task priority weights, and access overlap duration, and can more accurately reflect the conflict risk level of the node in the future scheduling cycle. In addition, to improve the flexibility of the scheduling system, real-time updates and dynamic queries of the path intersection density factor are supported to ensure that when the task status changes, the intersection node set can respond and update in a timely manner.

[0031] The "path intersection density factor" is a comprehensive indicator used to measure the intensity of task access overlap at path nodes within a specific time window, and is intended to accurately reflect the resource competition risk level at path intersection nodes. This factor comprehensively counts multiple factors, including the number of tasks accessing the same path node in the same time window, the priority weights of each task, the degree of overlap of access time periods, and conflicts in access directions, and calculates the conflict density level of the node in the current time period through weighted aggregation. Its main function is to rank the risks of all path intersection nodes, thereby helping the scheduling module identify high-density, high-priority conflict areas, and perform path avoidance, time staggering, or task rescheduling in advance. By introducing the "path intersection density factor", the scheduling system can dynamically perceive changes in resource pressure within the scheduling cycle, accurately locate and regulate path conflict hotspots, and is a key decision factor for improving path planning accuracy and scheduling stability.

[0032] A path conflict distribution map is generated based on the set of path intersection nodes and their corresponding access frequency and cross-density data. This map uses a unified time axis as the horizontal axis and path space nodes as the vertical axis to form a visual two-dimensional conflict mapping map, where high-density areas represent potential high-risk path resource contention points. The path conflict distribution map can not only serve as the input basis for the subsequent path loop detection engine and task rescheduling module, but can also be used to assist dispatchers in path risk monitoring, resource priority setting, and scheduling strategy adjustment. Furthermore, the map supports visual hot zone identification, conflict level labeling, and cross-intensity trend analysis, improving the risk identification capability and decision-making transparency of the scheduling system in complex dynamic environments.

[0033] This step enables unified temporal analysis of the access behaviors of multiple mobile storage and charging robots on path resources. By mapping each task's path pre-occupancy graph onto a unified timeline, it is possible to identify overlapping accesses to path nodes within each time period, construct a set of path intersection nodes, and calculate their access frequency and path intersection density, thereby generating a path conflict distribution map. This map reflects the intensity of resource competition among different path nodes in the future time period, helps identify areas of high conflict risk, and provides the scheduling system with precise path avoidance and priority adjustment references. It is a prerequisite for path decoupling and task rescheduling.

[0034] The resource dependency graph construction module constructs a resource access dependency graph between tasks based on the path conflict distribution graph and the set of path intersection nodes. It identifies whether there is a closed access dependency chain between multiple mobile storage and charging robots on the path nodes in the resource access dependency graph. If so, it extracts the task set that forms a logical loop structure and generates a loop identification set. To proactively identify and provide early warning of scheduling deadlock risks that may arise when multiple mobile storage and charging robots use path resources, we construct a resource access dependency graph between tasks based on the path conflict distribution graph and the set of path intersection nodes. Based on this structure, we identify closed dependency chains in path access, extract the set of tasks with potential deadlock risks, and output a loop identification set. The specific implementation process includes the following steps: Traverse all high-density path intersection nodes in the path conflict distribution graph, and combine the task access information in the path intersection node set to pair the access tasks of each path node and build edges, gradually constructing a resource access dependency graph between tasks. In this step, if task A and task B have a temporally overlapping access relationship at a certain intersection node, and based on the access order and task priority, it is determined that task A must wait for task B to release the path node, a directed edge is established between task A and task B, indicating that task A depends on task B to release the resource. The key to this step is to establish a task-level spatiotemporal dependency network with tasks as nodes in the graph and resource occupancy waiting relationships as edges, thereby abstracting potential path conflict relationships into a structured graph model, forming the basic data structure for deadlock determination in the scheduling system.

[0035] Perform loop identification on the directed graph of the constructed resource access dependency graph. Specifically, a closed-loop detection algorithm based on graph traversal is used. Starting from any node in the dependency graph, a recursive search is performed to see if there is a path that can return to the starting task node. If such a path exists, it means that a closed dependency chain has been formed between the tasks, that is, task A waits for task B, task B waits for task C, and finally task C indirectly waits for task A, forming a "waiting loop" in the scheduling logic. Once this loop structure is formed, the path node will be permanently locked, the robot task will not be able to move forward, and the resources will not be released. This is a typical deadlock trigger mode. The loop detection algorithm can efficiently identify such logical structures, providing an accurate basis for subsequent deadlock warning and scheduling decoupling.

[0036] A closed-loop detection algorithm based on graph traversal is used to identify loops in directed graphs. It is often used to determine whether there are circular waits in task or resource dependencies. The algorithm's core function is to structurally analyze the resource access dependency graph to determine whether multiple mobile storage and charging robots form closed dependency chains when accessing path nodes. This involves multiple tasks waiting for each other to release path resources, ultimately leading to a logical deadlock where tasks cannot progress. Starting from any task node in the dependency graph, the algorithm uses a depth-first traversal or topological sorting method to recursively check along the dependency edges layer by layer to see if it can return to the starting node. If so, a closed loop exists, indicating a "wait loop" between tasks. This algorithm boasts high recognition efficiency and structural versatility. It can quickly identify potential deadlock-risk paths in the system without interfering with scheduling execution, providing a foundational basis for subsequent actions such as task rescheduling and path decoupling. It is a key analytical tool for ensuring the operational stability and path security of scheduling systems.

[0037] After detecting the existence of one or more closed access dependency chains in the resource access dependency graph, all task numbers participating in the closed loop structure are extracted and formed into a "loop task set". At the same time, a unique identification information is generated for each closed loop structure, called a "loop identification". The identification includes fields such as the loop structure number, the list of tasks involved, the list of intersection nodes, the time window range of the conflict, and the path intersection density statistics, forming a structured data unit. The set of all loop identifications is uniformly stored as a "loop identification set" as a direct input parameter for the subsequent task decoupling and rescheduling mechanism. The key to this step is not only to identify the deadlock risk, but also to accurately mark the source, impact range and conflict intensity of the deadlock risk in a structured manner, so that the scheduling system has clear intervention positioning and analysis capabilities.

[0038] In order to ensure the dynamic responsiveness of the resource access dependency graph, a periodic scheduling mechanism and event-driven update mechanism for loop detection are introduced. When the scheduling system detects trigger conditions such as task status changes, path re-planning, task priority adjustments, or path node reachability updates, it will automatically rebuild the resource access dependency graph and re-execute the loop detection process to ensure that the loop identification set always reflects the current latest dependency status. At the same time, it supports exporting the loop identification set as a graphical report or using it for scheduling simulation to assist the scheduling optimization algorithm in pre-processing high-risk paths, adjusting the execution order, or setting access weights, and avoiding potential path conflict loops in advance, thereby improving the overall robustness of the scheduling system and task continuity.

[0039] This step aims to identify potential task deadlock risks arising from path resource sharing in the scheduling system. By constructing a resource access dependency graph between tasks based on a path conflict distribution graph and a set of path intersection nodes, we can abstract the waiting and dependency relationships between tasks at path nodes. Furthermore, we can detect the existence of closed access dependency chains, where multiple robot tasks are unable to progress due to waiting for path release. Once such loop structures are discovered, the corresponding task sets are extracted and a loop identification set is generated. This provides a clear basis for subsequent task decoupling, path rerouting, and deadlock avoidance, ensuring stable operation of the scheduling system and preventing congestion.

[0040] The decoupling and rescheduling module decouples and reschedules tasks based on loop identification sets, adjusting the task execution order or path node access time of some mobile storage and charging robots, and generating an updated task timing graph to break the closed dependency chain structure in the resource access dependency graph and avoid scheduling deadlocks caused by path node resource contention. To effectively break the closed loop structure formed by resource dependencies during the path access process of multiple mobile storage and charging robots and avoid scheduling deadlock caused by tasks waiting for each other to release path nodes, based on the loop identification set generated above, task decoupling and sequence reordering operations are performed to break the circular path in the dependency chain at the scheduling level, achieving the unlocking of task execution and smooth switching of resource access. This process includes the following steps: The identified loop task sets are extracted one by one from the loop identification set, and the tasks involved in each set are comprehensively evaluated based on multi-dimensional indicators such as task priority, path conflict density, node overlap, and task timeliness requirements. During the evaluation process, a "scheduling intervention impact factor model" is constructed. This model is used to measure whether the closed loop can be broken at the lowest cost after adjusting a specific task in the current loop structure, while keeping the overall execution efficiency as unaffected as possible. Based on the evaluation score output by the model, the key task that is most suitable as the "decoupling entry point" is identified. This task will be adjusted first in subsequent operations to drive the deconstruction of the entire closed-loop structure.

[0041] The "Scheduling Intervention Impact Factor Model" is a key evaluation mechanism used to guide task decoupling and rescheduling decisions. Its goal is to identify the optimal intervention path among multiple candidate scheduling intervention options—that is, to find the task adjustment solution that breaks the path dependency loop at the lowest cost while minimizing the impact on the overall task cadence. The model's construction involves the following steps: All tasks participating in the loop are extracted based on a loop identification set. For each task, its "criticality weight" within the loop structure is calculated. This weight is evaluated based on the task's position within the loop (e.g., whether it is a starting node or an intermediate node with a high proportion of shared paths). Each task's schedulability metrics, including task priority, time window slack, and the strength of its binding with path nodes, are analyzed to determine its adaptability and the cost of such an adjustment. The impact of each candidate task's adjustment on the overall scheduling cadence is evaluated, including its timing traction effect on other tasks, resource release delays, and changes in path resource utilization, to construct a "post-intervention effect function." The criticality weight, schedulability metrics, and post-intervention effect function are weighted and fused to output a normalized intervention impact score, forming a prioritized list of task decoupling options. The role of this model is to provide a quantitative decision-making basis for the scheduling system, helping to quickly select the optimal intervention point when faced with multiple possible decoupling paths, thereby efficiently breaking the path conflict loop, while taking into account the minimization of scheduling costs and the stability of task execution, ensuring that the scheduling optimization process is scientific, explainable and practical.

[0042] For the selected key tasks, a decoupling strategy path is determined, including adjusting the execution order of the tasks or modifying the access time of the path nodes. The execution order adjustment is mainly applicable to scenarios where the task can be reasonably delayed within the task plan without affecting the overall business logic; while the access time adjustment is mainly achieved by introducing a delay factor or a time peak processing mechanism to the access time period of the path node, so that the access windows of the tasks are staggered as much as possible to reduce the risk of conflict. In order to improve the accuracy and controllability of scheduling, the "time control weight coefficient" is introduced as an auxiliary decision-making factor to constrain the maximum offset range of the time adjustment, prevent the decoupling operation from causing the task completion time limit to deviate beyond the tolerance range, and ensure that the scheduling intervention is carried out within the safety boundary.

[0043] The "time control weight coefficient" is a control parameter introduced to constrain the scope of scheduling adjustments. Its core meaning is to set a tolerable time offset upper limit for each task, which measures the maximum range of power generation delays that the task can accept during decoupling operations. This coefficient is dynamically set based on a combination of factors, including the task's time sensitivity, priority level, resource dependency, and business completion deadline. A smaller value indicates a task's greater sensitivity to timing adjustments and a more limited margin for adjustment. Its purpose is to provide a "time safety margin" during task decoupling and path staggering adjustments, preventing excessive delays in certain critical tasks from impacting their normal completion or triggering cascading scheduling conflicts. In practice, when evaluating task adjustability, the difference between the planned task access time and the adjusted time is calculated, and the resulting offset is compared to the permissible range corresponding to the time control weight coefficient. A scheduling adjustment is considered "legal and acceptable" only if the offset does not exceed the preset time threshold. This mechanism ensures the controllability of decoupled operations and the consistency of task completion deadlines. It not only avoids a new round of task conflicts caused by forced peak shifting, but also improves the accuracy and robustness of the overall scheduling strategy. It is an important parameter means to achieve flexible scheduling and task timing safety control.

[0044] The adjusted task information is rewritten into the task scheduling table, and an "updated task sequence diagram" is regenerated accordingly. This task sequence diagram, based on the timeline, provides a structured representation of the task execution sequence and path node access plan for all mobile storage and charging robots, showcasing the changes in task rhythm and path occupancy for each robot throughout the scheduling cycle. Simultaneously, the task path pre-occupancy graph and resource access dependency graph are reconstructed to verify whether the adjusted task has successfully broken the original loop structure. If the corresponding closed-loop path in the dependency graph has been disconnected, the decoupling adjustment is effective; if not, a second round of intervention is performed by selecting a suboptimal task from the loop identification set until all loop structures are completely deconstructed.

[0045] To ensure that the scheduling changes after task decoupling and rescheduling are accurate and visually evaluated, the adjusted task information must be rewritten into the task schedule and an "updated task timing diagram" generated based on this information. The specific steps include: Based on the adjustment results output by the aforementioned task decoupling logic, each mobile storage robot's task execution order, path node access time, start and end times, and related priority information are structurally rewritten and updated into the global task schedule; The timing diagram construction module is called to read the key time fields of each task in the latest task schedule, group each task by robot number, and map the execution trajectory of each robot throughout the entire scheduling cycle according to a unified timeline sequence; Combining path node information and task time periods, a two-dimensional graph is drawn with "time" as the horizontal axis and "robot number or path node" as the vertical axis, marking the task start, end, waiting, conflict avoidance, and other state segments, thus constructing an intuitive task scheduling view; The original timing diagram is compared with the updated timing diagram to calculate the start and end offsets of each task and the impact range of the scheduling change. This graph is then used for subsequent scheduling performance evaluation and optimization strategy feedback. The core function of the "updated task timing diagram" is to provide a dynamic, visual, and clearly structured expression of task scheduling status, enabling the scheduling system to monitor the execution effect of scheduling adjustments in real time, determine whether loop dependencies have been completely broken, and verify whether decoupling operations are completed within the time safety boundary. At the same time, it provides an intuitive data basis and decision support for path priority adjustment and resource reallocation.

[0046] To enhance adaptability and scheduling intelligence in dynamic environments, the task decoupling module further introduces an "adaptive decoupling scheduling strategy set." This strategy set dynamically selects the most appropriate decoupling solution based on real-time indicators such as current load status, task urgency, robot availability, and path resource utilization. Adjustment records and loop releases are also archived to build a "scheduling optimization feedback library" for subsequent scheduling strategy training and optimization algorithm learning, enabling the continuous evolution of the scheduling system and improving its control capabilities.

[0047] The "Adaptive Decoupling Scheduling Strategy Set" is a set of configurable, switchable strategies for dynamic scheduling optimization in complex multi-robot task conflict scenarios. Its core concept is to automatically select the most appropriate decoupling solution for the current scheduling environment based on real-time indicators such as current operating status, task load, path conflict, and resource utilization, and dynamically adjust the scheduling strategy to maintain stability and efficiency. The strategy set includes various predefined decoupling logics, such as priority control, peak-shifting, path detour, and task suspension, each designed for specific loop structures and resource conflict scenarios. Its primary function is to provide the scheduling system with "scenario awareness" and "strategy adaptation" capabilities. This allows it to quickly select the optimal decoupling solution without manual intervention in situations such as high concurrency, severe path overlap, or complex task priority conflicts, thereby improving scheduling responsiveness and robustness. By introducing the "Adaptive Decoupling Scheduling Strategy Set," the scheduling process no longer relies on a single fixed strategy, but instead features flexible switching and self-adjustment capabilities, achieving intelligent, refined, and efficient scheduling control.

[0048] This step aims to proactively break resource dependency loops between tasks through scheduling optimization, avoiding scheduling deadlocks caused by resource contention at path nodes. Based on the loop identification set, task combinations with closed dependency chains are identified, and some critical tasks are decoupled and rescheduled, including adjusting the task execution order or path node access times to change the original conflict structure. The resulting updated task timing diagram not only eliminates waiting loops between tasks but also optimizes the task execution rhythm and path usage distribution, providing a stable, continuous, and efficient task execution sequence. This is a key step in proactively avoiding scheduling deadlocks.

[0049] The timing offset assessment module compares the updated task timing diagram with the original task time information, calculates the difference between the start and end time of each mobile storage and charging robot task, and outputs the timing offset data corresponding to each task to evaluate the impact of scheduling adjustments on the task execution rhythm; To comprehensively evaluate the impact of task decoupling and path rerouting on the execution rhythm of each mobile storage and charging robot, a task timing offset analysis process was implemented. By performing a structured comparison between the updated task timing diagram and the original task plan, the task time changes of each robot were accurately quantified. Standardized timing offset data was then output to provide feedback on the degree of rhythm disturbance caused by the scheduling adjustment. This implementation process includes the following steps: Based on the task schedule before task decoupling and rescheduling, the original start and end times of all tasks are extracted and grouped by task number and robot number to construct an "original task time mapping table." This mapping table serves as a reference template for offset calculation, preserving the basic structure of the overall task rhythm before decoupling adjustments and ensuring a clear time base for offset analysis. Simultaneously, the adjusted start and end times of all tasks are extracted from the updated task timing diagram, and an "adjusted task time mapping table" is constructed that matches the original table structure, enabling subsequent one-to-one comparisons at the task level.

[0050] A "start time difference calculation" and "end time difference calculation" are performed on each task to determine the advance or delay offset of the task's start time after decoupling and rearrangement, as well as the magnitude of the change in completion time. To improve the operability of the results, this time offset information is structured into a "task timing offset data tuple," which includes parameters such as task number, robot number, start time offset value, end time offset value, offset direction (advance or delay), and whether it exceeds the task category rhythm tolerance threshold. The task rhythm status is automatically labeled, such as "stable," "slight fluctuation," or "imbalance warning." This structural design not only facilitates the rapid identification of tasks significantly affected by scheduling changes, but also provides a clear basis for subsequent scheduling feedback mechanisms.

[0051] The timing offset data tuples of all tasks are aggregated to construct a "task timing offset matrix". The matrix has robot numbers as rows and task numbers as columns. The offset values ​​and status labels are filled in the matrix cells to form a visual rhythm change map. On this basis, it supports setting the offset severity threshold and highlighting tasks with large offset fluctuations, so that dispatchers can quickly identify rhythm-sensitive sections. At the same time, the present invention introduces a "task rhythm stability scoring" mechanism. By analyzing the total amount and fluctuation range of each robot's task offsets throughout the scheduling cycle, it comprehensively judges whether its overall execution rhythm is stable, providing decision support for path priority adjustment and task rescheduling.

[0052] The above analysis results are summarized to generate a "Scheduling Rhythm Impact Assessment Report," which includes the global average task offset, a list of tasks with the largest offset, a ranking of robot rhythm fluctuations, and a list of recommended intervention tasks. This report is used to evaluate the rhythm-friendliness and stability of the current scheduling adjustment strategy. The assessment results are stored in the scheduling history database, serving as reference data for subsequent scheduling optimization model training and can also be used to compare scheduling strategy versions and predict trends.

[0053] This step accurately assesses the impact of scheduling adjustments (such as task decoupling and path rerouting) on ​​the execution rhythm of each mobile storage and charging robot's tasks. By comparing the updated task timing diagram with the original task time plan, the change in the start and end time of each task is calculated and standardized timing offset data is output. This offset data can be used to determine whether the scheduling adjustment has caused task rhythm imbalance, execution delays, or the spread of resource conflicts. This provides a quantitative basis for scheduling strategy optimization, path priority adjustment, and task timing control, ensuring that the scheduler maintains overall rhythm stability and time coordination while completing tasks efficiently.

[0054] The path weight control module combines the path conflict distribution map and timing offset results to construct a path control weight matrix. Based on the conflict density of path nodes and the offset degree of task timing, it dynamically adjusts the access priority and release time of path nodes. During the path selection process, it guides the robot to prioritize avoiding path nodes with high conflict, large offset, and concentrated synchronization pressure, thereby achieving continuous improvement in path decoupling and task scheduling stability. To achieve closed-loop control of scheduling optimization and dynamic regulation of path planning, based on the path conflict distribution map and task timing offset results, a "path control weight matrix" is further constructed to dynamically adjust the access priority and release timing of path nodes, guiding the robot to avoid conflict-intensive areas and rhythm-unbalanced paths, thereby achieving effective decoupling of path resources and continuous improvement of task scheduling stability. This process includes the following steps: Data is extracted from all path intersection nodes in the path conflict distribution map. The conflict density value for each node within the scheduling cycle is extracted and normalized with the path intersection density factor to form a "node conflict weight vector." Simultaneously, the task timing offset matrix is ​​used to aggregate and weight the timing offset values ​​of the tasks associated with each path node within the corresponding time window to obtain a "node rhythm pressure index," which measures the degree of synchronization pressure on the node during a specific period. By combining the data from the two dimensions of conflict density and rhythm pressure, a basic weight data item is generated for each path node, which serves as the original building block of the path control weight matrix.

[0055] Based on the basic weight data items of each path node, a "path node priority adjustment rule" is introduced to set the current access priority level and release time control factor for each node. The priority level is a numerical graded representation of the degree to which a node is preferentially accessed or avoided during the scheduling process. The release time control factor defines the robot's minimum residence time and necessary waiting time at the node, ensuring the temporal safety of path access. To enhance control flexibility, a "conflict-offset joint adjustment coefficient" is introduced. Based on the product of conflict density and offset amplitude, the access strategy of each node is dynamically determined, enabling it to make balanced path allocation decisions under the dual pressures of conflict and rhythm imbalance.

[0056] The generated priorities and release timing parameters are aggregated to construct a complete path control weight matrix. This matrix, indexed by path nodes, records multi-dimensional parameters such as access priority, release timing, and cadence influencing factors, segmented by scheduling time windows. This matrix is ​​used to drive the dynamic path selection process of the path planning engine in real time. During the task path planning phase, when the robot faces multiple candidate path nodes, it automatically references the weight data for the current time period in the weight matrix to guide the robot to select a path with low conflict, minimal cadence impact, and manageable synchronization pressure, proactively avoiding potentially risky nodes and optimizing resource usage distribution.

[0057] To enhance continuous control capabilities, the path control weight matrix is ​​designed as an updateable dynamic structure. When path conflict status or task cadence data changes significantly, the weight parameters of the relevant nodes are recalculated and the matrix content is updated, forming a "real-time feedback mechanism for path control." Furthermore, historical weight adjustment records are archived for use in training scheduling models, optimizing path guidance strategies, and enabling predictive modeling of future path conflict trends, further enhancing scheduling adaptability and operational stability in complex dynamic environments.

[0058] This step aims to construct a path control weight matrix, integrating path conflict density with task timing offset results, and dynamically adjusting the access priority and release timing of path nodes. This allows the mobile storage and charging robot to proactively avoid path nodes with high conflict, large offset, and concentrated synchronization pressure during path planning. This mechanism effectively alleviates path resource contention, reduces task execution interference, and prevents the formation of new scheduling loops. It achieves path-level task decoupling and continuous optimization of the cadence, ensuring the stability, flexibility, and continuous control capabilities of the scheduling system in a multi-task, high-concurrency environment.

[0059] By introducing a multi-level dynamic scheduling mechanism involving path dependency modeling, path conflict prediction, resource dependency graph construction, and task decoupling and rescheduling, this invention enables the early identification, precise location, and effective intervention of path conflicts and scheduling deadlock risks faced by mobile storage and charging robots in complex environments, significantly improving scheduling robustness and execution continuity. Compared to existing technologies, this solution not only possesses high-precision resource conflict identification capabilities for asynchronous multi-robot scheduling, but also adaptively optimizes path priority and access timing by combining timing offset assessment and path control weights, thereby avoiding deadlocks caused by resource loop dependencies, ensuring smooth progress of robot tasks and timely completion of energy replenishment, and improving overall operational efficiency and scheduling stability.

[0060] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0061] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

[0062] It should be noted that, in this document, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0063] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0064] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0065] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0066] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0067] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0068] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0069] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

Claims

1. A remote dispatching and path planning system for mobile storage and charging robots based on the Internet of Things, characterized by: It includes path dependency modeling module, path conflict analysis module, resource dependency graph construction module, decoupling and rescheduling module, timing offset evaluation module and path weight control module: The path dependency modeling module builds a task path dependency extraction model, obtains the path node sequence and access time period of each mobile storage and charging robot's current task to be executed, and generates a task path pre-occupancy map; The path conflict analysis module maps the pre-occupancy graphs of each task path onto a unified time axis, identifies the access overlap of path nodes in each time period, constructs a set of path intersection nodes, calculates the path intersection density, and generates a path conflict distribution map. The resource dependency graph construction module builds a resource access dependency graph between tasks based on the path conflict distribution graph and the path intersection node set, identifies whether there is a closed access dependency chain, and if so, extracts the task set and generates a loop identification set; The decoupling and rescheduling module decouples and reschedules tasks based on the loop identification set, adjusts the execution order of some mobile storage and charging robots or modifies the path access time, generates an updated task timing diagram, and verifies whether the loop structure is broken; The timing offset evaluation module compares the updated task timing diagram with the original task time information, calculates the difference between the start and end time of each task, and outputs the task timing offset data; The path weight control module combines the path conflict density and task timing offset results to construct a path control weight matrix, dynamically adjusts the access priority and release time of path nodes, and guides path selection to avoid high-conflict and high-offset nodes.

2. The remote scheduling and path planning system for mobile storage and charging robots based on the Internet of Things according to claim 1 is characterized in that: Building a task path dependency extraction model includes the following steps: Obtain the starting position, target position, expected execution time window and task priority of each mobile storage and charging robot's current task; Call the path planning engine to generate a path that includes a sequence of path nodes and the entry and exit times of each node; The path nodes and their time information are constructed as node time label tuples, which include path node identification, entry timestamp, exit timestamp and task priority factor; All node time label tuples are aggregated into a task path preemption set and uniformly constructed into a task path preemption graph.

3. The remote scheduling and path planning system for mobile storage and charging robots based on the Internet of Things according to claim 1 is characterized in that: Mapping the task path preemption graph to a unified timeline and generating a path conflict distribution graph includes the following steps: Perform temporal reordering on the entry and exit timestamps in each node's time label tuple to complete unified timeline mapping. In each time window, the path nodes visited by multiple tasks are counted and cross-feature verification is performed to determine the path intersection nodes; Construct a set of path intersection nodes and calculate the visit frequency and path intersection density factor of each node; A path conflict distribution graph with time axis and space nodes as coordinates is generated based on the path intersection node set.

4. The remote scheduling and path planning system for mobile storage and charging robots based on the Internet of Things according to claim 1 is characterized in that: Building a resource access dependency graph and generating a loop identification set includes the following steps: Based on the path conflict distribution graph and the path intersection node set, For task pairs with overlapping access relationships and waiting dependencies, a resource access dependency graph is constructed with tasks as nodes and dependency relationships as edges. Identify graph traversal loops on the dependency graph to detect whether there is a closed access dependency chain; If there is a closed dependency chain, the corresponding task number, intersection node, and conflict time period are extracted to generate a loop identifier; All loop identifications are aggregated into a loop identification set for subsequent scheduling optimization processing.

5. The remote scheduling and path planning system for mobile storage and charging robots based on the Internet of Things according to claim 1 is characterized in that: Task decoupling and rearrangement based on the loop identification set includes the following steps: Extract the loop task set from the loop identification set, evaluate the impact of each task on the decoupling effect based on the scheduling intervention influence factor model, and determine the key tasks; Adjust the execution order of key tasks or modify the path access time, and introduce a time control weight coefficient to limit the maximum offset range; Write the adjustment results into the task scheduling table, generate an updated task timing diagram, and reconstruct the task path pre-occupancy diagram and resource access dependency diagram; If the loop structure is not completely eliminated, the decoupling process is executed repeatedly until the loop is eliminated, and the decoupling scheme is dynamically optimized using the adaptive decoupling scheduling strategy set.

6. The remote dispatching and path planning system for mobile storage and charging robots based on the Internet of Things according to claim 1 is characterized in that: Outputting the timing offset data corresponding to each task includes the following steps: Extract the start and end times of tasks before and after decoupling, and construct the original task time mapping table and the adjusted task time mapping table; Calculate the offset between the start and end time of each task, and generate a task timing offset data tuple containing the offset value, offset direction and rhythm status; Aggregate all timing offset data into a task timing offset matrix and set the offset severity threshold to highlight it; Generate a scheduling rhythm impact assessment report and store the results in the scheduling history database for subsequent strategy optimization.

7. The remote scheduling and path planning system for mobile storage and charging robots based on the Internet of Things according to claim 1 is characterized in that: Constructing the path control weight matrix includes the following steps: Extract the path node conflict density in the path conflict distribution map and the offset data in the task timing offset matrix; According to the conflict density of each path node and the timing deviation degree of the associated tasks, the access priority level and release time parameters are set; Generate a path control weight matrix with path nodes as indexes, and record the priority and release sequence of each node in different time windows; During the path selection process, the weight matrix is ​​referenced to dynamically guide the robot to avoid high-conflict and high-offset path nodes.

Images