Battery charging time planning system based on cycling behavior analysis

Abstract

This invention relates to the field of battery charging planning technology, specifically a battery charging time planning system based on cycling behavior analysis. The system includes an inertial recognition module, a density recognition module, a node adjustment module, a buffer discrimination module, and a time period filtering module. This invention deeply analyzes the power change signals in robot operation data, fusing inertial response and acceleration characteristics. Based on the multi-level characteristics of task segments in time and space, it identifies power nodes closely related to path behavior. Utilizing segment distribution density, energy consumption variation ranges, and task load status, it comprehensively regulates the determination of energy consumption nodes within each task segment, achieving intelligent linkage between task rhythm, energy consumption buffer, and actual available charging range. Through dynamic screening of behavioral inflection points and task gaps, it forms an adaptive match between charging time periods and task flow, reducing resource conflicts and improving the continuity and response efficiency of charging scheduling in multi-task environments.

Description

Technical Field

[0001] This invention relates to the field of battery charging planning technology, and more particularly to a battery charging time planning system based on cycling behavior analysis. Background Technology

[0002] Battery charging planning primarily involves modeling and scheduling battery energy usage and replenishment behavior in different application scenarios. This field is particularly crucial in autonomous operating platforms such as robots. Battery charging planning technology also covers aspects such as charging station sharing and queuing strategies, charging priority strategies, and battery health status assessment in multi-robot systems. Traditional battery charging time planning systems utilize operational behavior data exhibited by mobile platforms during task execution, such as path trajectories, speed changes, task frequency, and energy consumption rates, to estimate power consumption through behavioral modeling and predict future charging needs. This approach focuses on how to combine the behavioral characteristics of robot movement to achieve advance planning of battery charging time.

[0003] Existing technologies rely on basic robot motion data and energy consumption changes to make unified judgments in path planning and battery charging time management. However, they lack in-depth analysis of the dynamic path inertial characteristics and task segment rhythm density. In actual multi-task operation scenarios, it is difficult to detect energy consumption anomalies and task connection contradictions caused by path behavior fluctuations in a timely manner. Especially in environments with repetitive paths, frequent changes in task segments, or high-load tasks, problems such as disconnection between charging node arrangement and actual operation, poor connection between task segments, or decreased resource allocation efficiency may occur, which restricts the intelligence and adaptability of charging scheduling systems in complex scenarios. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a battery charging time planning system based on cycling behavior analysis.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a battery charging time planning system based on cycling behavior analysis, the system comprising:

[0006] The inertial recognition module analyzes historical paths based on robot operation information, compares trajectory differences by combining load changes, identifies sudden motion changes and continuous segments of the path, and calibrates node changes based on navigation data to obtain path dynamic feature nodes.

[0007] Based on the path dynamic feature nodes, the density recognition module determines the start and end nodes of each task segment, analyzes the temporal connection features between task segments, compares the time span differences between segments, identifies dense and scattered task intervals, and obtains the task segment distribution features.

[0008] Based on the distribution characteristics of the task segments, the node adjustment module analyzes the changes in the remaining power and output power within the task segments, determines the synchronization relationship between rhythm and energy consumption, adjusts the charging start judgment criteria, and obtains the energy consumption dynamic control node.

[0009] The buffer discrimination module analyzes the energy consumption change characteristics of task segments based on the energy consumption dynamic control node, determines the relationship between energy consumption and running status, compares energy consumption distribution, identifies energy consumption buffer segments, and obtains energy consumption buffer markers.

[0010] Based on the energy consumption buffer zone markers, the time period filtering module analyzes the distribution of repeated path segments, compares the coverage of available charging segments in the scheduling sequence, identifies non-overlapping charging intervals, and obtains the available charging scheduling intervals.

[0011] The present invention improves upon the following: the path dynamic feature nodes include motion response attributes, node distribution status, and trajectory change type; the task segment distribution features include distribution interval attributes, segment continuity characteristics, and task density; the energy consumption dynamic control nodes include control trigger time, node corresponding energy consumption level, and control reference parameters; the energy consumption buffer zone markers include buffer identification segments, energy consumption change intervals, and task status labels; and the charging scheduling available interval includes charging allocation time periods, task interval blocks, and available scheduling labels.

[0012] The present invention is improved in that the inertial recognition module includes:

[0013] The data stream receiving submodule analyzes the attitude and angular velocity information collected by the inertial measurement device based on the robot's operation information, combines the acceleration and velocity information recorded by the accelerometer, compares the changing trends of various data within the same time period, identifies the time segments associated with changes in path direction, and obtains the inertial response segments.

[0014] The path load comparison submodule compares the trajectory data under various load conditions based on the inertial response segment, analyzes the differences in path direction and orientation changes within each load interval, identifies sections with significant changes in angular velocity, determines the trajectory offset characteristics within each interval, and obtains the path offset pattern.

[0015] The node evolution extraction submodule analyzes motion mutation segments and continuous trajectory segments based on the path offset pattern, determines the positional changes of path nodes in each time period, compares the temporal distribution characteristics between nodes, identifies representative dynamic migration trajectories, and obtains path dynamic feature nodes.

[0016] The present invention is improved in that the density recognition module includes:

[0017] The start and end node identification submodule, based on the path dynamic feature nodes, determines the arrangement of adjacent nodes in the navigation sequence and their corresponding task status, filters the node combinations with task boundary features in the navigation sequence, identifies the start and end nodes of each task segment, and obtains the start and end nodes of the task segment.

[0018] The time connection structure analysis submodule analyzes the time connection characteristics between consecutive task segments based on the start and end nodes of the task segments, determines the execution order and degree of connection within and outside the task segments, identifies connection segments with different time spans, and obtains the task time connection sequence.

[0019] The task density judgment submodule determines the distribution range and quantity of task segments within each time interval based on the task time connection sequence, compares the continuity and dense distribution of adjacent task segments, identifies time intervals where tasks are frequent or scattered, and adjusts the criteria for dividing task intervals to obtain the distribution characteristics of task segments.

[0020] The present invention is improved in that the node adjustment module includes:

[0021] The remaining power monitoring submodule analyzes the time series and remaining power data of the task segments based on the distribution characteristics of the task segments, calculates the rate of power change between adjacent sampling points, determines the correspondence between power change and time process in each task segment, optimizes the power data collection method, and obtains the power change rate sequence.

[0022] The power trend comparison submodule compares the power change rate sequence with the sampled data of the output power of the task segment, analyzes the synchronization characteristics of power change and power change rate at each moment, and obtains the synchronization difference index.

[0023] The trigger node extraction submodule, based on the synchronization difference index, filters out time periods exhibiting synchronization anomalies, determines the task identifier and power change characteristics within the corresponding segment, and jointly analyzes the changes in the rate of power decrease to obtain the energy consumption dynamic control node.

[0024] The present invention is improved in that the buffer discrimination module includes:

[0025] The energy consumption trajectory extraction submodule analyzes the battery energy changes and time series of each task segment based on the energy consumption dynamic control node, calculates the fluctuation range of energy consumption in continuous time periods during operation, optimizes the stability of energy consumption data, identifies the energy consumption fluctuation characteristics under the operating state, and judges the influence of abnormal fluctuation signals to obtain the energy consumption fluctuation interval sequence.

[0026] Based on the energy consumption fluctuation interval sequence, the state mapping analysis submodule calculates the acceleration and velocity parameters of each task segment, compares the synchronization relationship between energy consumption and state characteristics under each operating state, judges the synergy between energy consumption change trend and state characteristics, determines the time period with prominent synchronization, and obtains the state synchronization interval sequence.

[0027] The buffer segment calibration submodule, based on the state synchronization interval sequence, determines the persistence characteristics of energy consumption fluctuations, compares the coupling offset between energy consumption and motion state within each time segment, and uses the formula:

[0028] ;

[0029] The coupling offset magnitude is obtained, and segments where the coupling offset magnitude converges continuously are identified to obtain the energy consumption buffer markers, where, Indicates the sequence number is The coupling offset magnitude of the task segment, Representing a task fragment The Middle Energy consumption rate per unit time period at each time point Representing a task fragment The Middle Acceleration parameters at each time point, Representing a task fragment The Middle Speed ​​parameters at each time point This represents the equivalent mass parameter, used to convert dynamic disturbance terms to the energy consumption level. Representing a task fragment Number of time nodes within.

[0030] The present invention is improved in that the time period filtering module includes:

[0031] The task gap identification submodule analyzes the sequential order of consecutive task segments in the time series based on the energy consumption buffer interval markers, determines the gaps between each task segment, filters time segments with interval characteristics, and obtains a set of task interval time segments.

[0032] The path segment localization submodule determines the distribution of path coordinates during each interval based on the task interval time segment set, analyzes the frequency of occurrence of trajectory segments in each task segment, identifies independently distributed and non-repeating trajectory segments, and excludes path segments with overlap to obtain an exclusive segment sequence.

[0033] The available interval filtering submodule compares the positional relationship between each candidate charging segment and the time interval of the scheduling sequence task based on the exclusive segment sequence, analyzes the overlap between the candidate segments and the task time interval, identifies the charging available segments that do not intersect, optimizes the scheduling interval rules, and obtains the charging scheduling available interval.

[0034] The present invention is improved in that the historical path refers to the path trajectory traversed by the robot during operation or experimentation, including a sequence of coordinate points and a time sequence; each task segment refers to a continuous motion interval of the robot on the path that has independent execution significance; and the start and end nodes refer to the positions on the path or task sequence at the beginning and end of each task segment.

[0035] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0036] In this invention, by deeply analyzing the dynamic change signals in the robot's operation data, the inertial response and acceleration characteristics are fused. Based on the multi-level characteristics of task segments in time and space, dynamic nodes closely related to path behavior are identified. By utilizing the segment distribution density, energy consumption variation range, and task load status, the energy consumption node determination within each task segment is comprehensively regulated, realizing intelligent linkage between task rhythm, energy consumption buffer, and actual charging available range. Through dynamic screening of behavioral inflection points and task gaps, an adaptive matching between charging time periods and task flow is formed, reducing resource conflicts and improving the continuity and response efficiency of charging scheduling in a multi-task environment. Attached Figure Description

[0037] Figure 1 This is a system flowchart of the present invention;

[0038] Figure 2 This is a flowchart of the inertial recognition module in this invention;

[0039] Figure 3 This is a flowchart of the density recognition module in this invention;

[0040] Figure 4 This is a flowchart of the node adjustment module in this invention;

[0041] Figure 5 This is a flowchart of the buffer discrimination module in this invention;

[0042] Figure 6 This is a flowchart of the time period filtering module in this invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0045] All user-related information involved in this invention (including but not limited to biometric information, identity verification information, behavioral data, device information, and other data that can be used for identity verification and personalized services) is collected and processed with the user's full knowledge and voluntary consent. The collection, storage, and use of all information strictly comply with applicable national and regional laws and regulations, and meet relevant data protection standards and policy requirements. The use of data is limited to purposes necessary for providing the technical services of this invention, and reasonable technical and management measures will be taken to ensure the security and confidentiality of users' personal information in terms of information protection and privacy.

[0046] Example

[0047] Please see Figure 1 This invention provides a technical solution: a battery charging time planning system based on cycling behavior analysis includes:

[0048] Based on the robot's operation information, the inertial recognition module analyzes the robot's historical operation path, correlates the motion characteristics collected by the inertial measurement device with the data recorded by the accelerometer, adjusts the acquisition action of each cycle of directional change, compares the differences in motion trajectory under load changes, identifies segments that correspond to motion abrupt changes and path continuity features, and obtains path dynamic feature nodes by comparing node evolution based on navigation data.

[0049] The density recognition module, based on path dynamic feature nodes, determines the start and end nodes of each task segment on the timeline, analyzes the temporal connectivity features inside and outside each task segment, compares the time span differences between consecutive task segments, identifies the dense or scattered features of task segments, adjusts the task division criteria, optimizes the task management interval, and obtains the task segment distribution characteristics.

[0050] Based on the distribution characteristics of task segments, the node adjustment module analyzes the remaining power collection information and output power change characteristics in each task segment, judges the synchronization degree between rhythm change and energy consumption trend, adjusts the charging start judgment criteria, compares the remaining power change with the current power trend, optimizes the triggering conditions of each segment, identifies nodes that can be used for judgment in the task segment, and obtains the energy consumption dynamic control nodes.

[0051] The buffer discrimination module analyzes the continuous energy consumption change characteristics of each task segment of the robot based on the energy consumption dynamic control node, judges the relationship between the energy consumption of each task segment and the running state, compares the energy consumption distribution during the operation, identifies the task segments with energy consumption change buffers, adjusts the scheduling judgment criteria, and obtains the marking between energy consumption buffers.

[0052] The time period filtering module determines the actual gap between robot tasks based on the energy consumption buffer zone markers, analyzes the distribution of repeated path segments in the time series, compares the coverage status of each available charging segment in the scheduling sequence, identifies non-overlapping time periods, optimizes scheduling rules, and obtains the available charging scheduling interval.

[0053] The path dynamic feature nodes include motion response attributes, node distribution status, and trajectory change type. The task segment distribution features include distribution interval attributes, segment coherence characteristics, and task density. The energy consumption dynamic control nodes include control trigger time, node corresponding energy consumption level, and control reference parameters. The energy consumption buffer zone markers include buffer identification segments, energy consumption change intervals, and task status labels. The available charging scheduling intervals include charging allocation time periods, task interval blocks, and available scheduling labels.

[0054] In the inertial recognition module, the historical running path refers to all the paths the robot has traversed during actual operations or experiments, including a sequence of coordinate points and a time sequence, which forms the basis of the riding behavior data; the collected motion features refer to the data related to the robot's posture, acceleration, angular velocity, and other physical motion states recorded by the inertial measurement unit; the recorded data refers to all raw data collected in real time by sensors such as accelerometers, including acceleration, velocity, and rate of change of acceleration; the correlation judgment refers to the joint analysis of the motion features collected by the inertial stabilizer and the accelerometer data to examine whether there are motion patterns or events between them. Consistency or causal relationship; acquisition of direction change refers to real-time data acquisition of direction change (such as gyroscope angular velocity) when the robot changes path or turns; motion trajectory difference refers to the changes in the robot's motion trajectory in the spatial or temporal dimensions under different loads, working conditions or paths; corresponding segments refer to specific time segments identified through data analysis that have a one-to-one correspondence or close correlation between motion characteristics and path changes; node evolution refers to the dynamic changes and migration patterns of key path nodes (such as inflection points, turning points, slope start and end points, etc.) in time and space as the robot runs.

[0055] In the density recognition module, each task segment refers to a continuous motion interval with independent execution significance that is divided on the robot's path, such as each delivery segment, each turn, or each straight-line movement segment; start and end nodes refer to the specific positions (time stamps or spatial points) on the path or task sequence at the beginning and end of each task segment; temporal connection characteristics refer to the temporal distribution characteristics of the connection, transition, and interval between task segments (such as whether they are continuous, the length of the interval, etc.); continuous task segments refer to multiple task segments that are closely connected or executed continuously in a predetermined order during task scheduling or path planning; time span differences refer to the changes or distribution of different task segments in duration, reflecting the dynamic characteristics of task load and intermittency; the dense or dispersed characteristics of task segments refer to the dense (short interval, frequent) or dispersed (long interval, sparse) characteristics of task segments distributed on the time axis or spatial path; task division criteria refer to the criteria for segmenting path tasks based on path inertia analysis, task triggering conditions, behavioral events, etc.; task management interval refers to the set of task segments used for system management, scheduling, and analysis, and their temporal and spatial intervals.

[0056] In the node adjustment module, the remaining power acquisition information refers to the monitoring data of the remaining available power fed back in real time by the battery management system; the output power change characteristics refer to the dynamic data of the robot's battery discharge, drive motor power output, etc., changing over time in different task segments; the rhythm change refers to the rhythm of changes in task intensity or behavior pattern during task execution, such as continuous high load or intermittent operation; the energy consumption trend refers to the direction of change in the robot's overall energy consumption within continuous task segments (such as gradual decrease, phased increase, etc.); the charging start judgment criteria refer to the conditions for determining when charging should be started, combining multiple indicators such as remaining power, task density, and power output; the trigger conditions for each segment refer to the judgment criteria or signals used when judging charging demand individually for each task segment (such as a sudden increase in energy consumption, etc.); the node used for judgment refers to the specific task interval determined as the charging start or energy consumption inflection point by analyzing task and energy consumption data.

[0057] In the buffer discrimination module, the continuous energy consumption change characteristic refers to the changing trend and fluctuation characteristics of the energy consumption data curve during the continuous operation of the robot; the relationship with the operating state refers to the inherent correspondence between energy consumption changes and the robot's working state (such as acceleration, deceleration, stopping, and standby); the energy consumption distribution during operation refers to the distribution pattern of energy consumption on the time axis during different time periods throughout the entire task execution period; the energy consumption change buffer refers to the section in the energy consumption curve where the change slows down or the fluctuation weakens, which is usually a suitable period for charging scheduling; the scheduling judgment standard refers to the system judgment rules and parameter set used when scheduling tasks and charging for the energy consumption buffer section.

[0058] In the time slot filtering module, actual gaps refer to the actual time gaps between different robot tasks or path segments, i.e., the intervals between non-continuous tasks; repeated path segments refer to segments where the robot passes through the same path multiple times in different tasks, used to analyze task density and path utilization efficiency; coverage status in the scheduling table refers to the distribution and overlap of all candidate charging time slots in the scheduling table, reflecting the coordination between charging demand and resource allocation; non-overlapping time slots refer to available charging time segments that do not overlap with other tasks or charging demands after system scheduling; scheduling rules refer to the strategies and priority settings used to filter and allocate charging time slots, including systematic arrangement rules such as avoiding resource conflicts and prioritizing critical tasks.

[0059] Please see Figure 2 The inertial recognition module includes:

[0060] The data stream receiving submodule analyzes the attitude and angular velocity information collected by the inertial measurement device based on the robot's operation information, combines the acceleration and velocity information recorded by the accelerometer, compares the changing trends of various data within the same time period, identifies the time segments associated with changes in path direction, and obtains the inertial response segments.

[0061] The robot acquires attitude data from the inertial measurement unit (IMU) in real time, and simultaneously acquires angular velocity information. The attitude data includes pitch, roll, and yaw angle changes to reflect the robot's real-time attitude adjustments in space. Angular velocity information comes from a three-axis gyroscope, reflecting the robot's rotational speed changes along each axis per unit time. Accelerometers then record the acceleration changes along the three axes. Combined with the robot's speed information while traveling along a specific path, the speed value is calculated using encoder increments read from the control module. After completing the above data acquisition, the four types of data are aligned according to the same timestamp to form a unified time-series data matrix. Each time segment is set to a length of 0.5 seconds. Within each time segment, the attitude change range, angular velocity fluctuation amplitude, acceleration fluctuation, and speed change value are detected. For each dimension of change... A threshold is set for each amplitude. If any two parameters exceed the threshold simultaneously, the segment is marked as a potential direction change segment. For example, during the robot's travel time interval from 18.0s to 18.5s, the pitch angle changes from 5° to 20°, the angular velocity on the z-axis jumps from 2° / s to 30° / s, and the speed drops sharply from 1.5m / s to 0.7m / s. This segment is identified as a suspicious inertial change segment due to a significant direction change signal. This segment is then mapped to the path data. Based on the spatial path coordinates of this time period, it is determined whether the path direction has changed by more than 30° in a short period of time. If a valid change in direction is confirmed, the time period is officially marked as an inertial response segment. For example, in the above case, the path suddenly changes from due east to north, with a direction change angle of 45°. This time period is ultimately confirmed as a valid inertial response segment.

[0062] The path load comparison submodule is based on inertial response segments. It compares trajectory data under various load conditions, analyzes the differences in path direction and orientation changes within each load interval, identifies sections with significant changes in angular velocity, judges the trajectory offset characteristics within each interval, and obtains the path offset pattern.

[0063] Based on the pre-set load conditions, the segments were divided into corresponding load categories: 0kg, 15kg, and 30kg. Each load corresponds to a set of trajectory data, derived from the robot's actual running path records under different task conditions. The three types of trajectories were segmented by time period, and the spatial path point sequences corresponding to the inertial response segments were extracted. The changes in path morphology were then compared. In the unloaded state, the trajectory adhered more closely to the set route when turning, while in the 30kg state, the turning path showed an outward expansion trend, with a significant increase in spatial offset. Simultaneously, angular velocity data under each load condition were compared. In the unloaded state, the angular velocity changed more rapidly, with shorter and more concentrated change segments, while in the 30kg load state... The angular velocity increase is slower but lasts longer. During the comparison, the section where the angular velocity change exceeds the benchmark value is identified as the section with significant change. For example, under a 15kg load, the angular velocity increases from 5° / s to 28° / s in 3 seconds, with a duration of 2.5 seconds. If this combination of change magnitude and duration exceeds the set benchmark condition, this section is identified as a section with significant angular velocity change. Further comparison of the path endpoint offset shows that the deviation between the endpoint and the target point is less than 0.2m under 0kg conditions, but exceeds 0.6m under 30kg conditions, indicating a clear path deviation trend. Based on this, it is determined that this section has trajectory deviation characteristics, and the path segment is marked as a path deviation pattern sample, with its occurrence location and time interval recorded.

[0064] The node evolution extraction submodule analyzes motion mutation segments and continuous trajectory segments based on path offset patterns, determines the positional changes of path nodes in different time periods, compares the temporal distribution characteristics between nodes, identifies representative dynamic migration trajectories, and obtains path dynamic feature nodes.

[0065] The complete path trajectory within 5 seconds before and after each offset path segment is extracted to form an analysis window. Within this window, points where the path direction changes every second are extracted as path nodes. These nodes are then arranged sequentially according to their appearance time, forming a time-series node chain. Continuous analysis is performed on the time intervals between nodes. By statistically analyzing the time differences between all adjacent nodes, the temporal density of node distribution is identified. For example, if multiple nodes are evenly distributed and their time differences are stable within 0.3 seconds, the segment is considered a continuous node segment. If the time intervals between nodes fluctuate drastically, with some intervals exceeding 1.2 seconds and the immediately following interval falling back to within 0.4 seconds, then a sudden change is identified in the segment. Further analysis is then performed based on this. Analyze spatial displacement changes and calculate the distance changes between nodes in the path graph. If the distance between two nodes suddenly increases by more than 1.0m and occurs during the abrupt change time period, the segment is marked as a representative trajectory abrupt change segment. For example, during robot transportation, if a path segment moves 1.5m from node A to node B within 2 seconds with a time interval of 0.6s, and then moves only 0.2m to node C with a time interval of 1.3s, then node B is the inflection point of change, and the trajectory is identified as a dynamic migration trajectory segment. Select the three nodes with the largest directional change angle and the shortest time interval from this trajectory segment as path dynamic feature nodes and record them for subsequent task segment energy consumption analysis and control node identification.

[0066] Please see Figure 3 The density recognition module includes:

[0067] The start and end node identification submodule is based on path dynamic feature nodes. It determines the arrangement of adjacent nodes in the navigation sequence and their corresponding task status, filters the node combinations with task boundary features in the navigation sequence, identifies the start and end nodes of each task segment, and obtains the start and end nodes of the task segment.

[0068] Extract all identified path node information from the robot navigation sequence. Each node contains three types of data: spatial coordinates, timestamp, and task identifier. Then, arrange the nodes in chronological order to construct a navigation sequence table. Next, sequentially call each pair of adjacent nodes to determine if there are task switching characteristics between the two nodes. The action is determined by reading whether the task identifier field has changed, and by combining this with whether the path segment to which the node belongs crosses a new scheduling task instruction. If the task identifiers of two nodes are different, and their interval is greater than the set minimum task switching interval (set to 10 seconds), then the pair of nodes is recorded as a candidate task boundary node combination. Examine the spatial overlap of all candidate node combinations in the path graph. If a path endpoint is found... If the spatial distance between the coordinates and the starting point of the next task is less than 1.5 meters, it is considered to satisfy the task connection logic of physical continuity. This combination is retained as a valid set of start and end nodes. Then, the time span of the node combination is further screened to exclude task segments with a duration of less than 3 seconds or more than 1800 seconds. The node combination that passes the spatial continuity judgment and time span screening is output as the start and end nodes of the task segment. For example, in a delivery task, the robot arrives at node A at 12:00:05 and then starts the next segment at 12:00:20. The overlap between the starting point of this segment and the end point of the previous segment is 1.2 meters. The two task identification numbers are different and the interval time is 15 seconds. Then, this point pair is recorded as a set of task segment start and end nodes.

[0069] The temporal connection structure analysis submodule analyzes the temporal connection characteristics between consecutive task segments based on the start and end nodes of the task segments, determines the execution order and the tightness of connection within and outside the task segments, identifies connection segments with different time spans, and obtains the task temporal connection sequence.

[0070] The start and end times of each task segment are extracted sequentially and sorted in ascending order. The time interval between two consecutive task segments is then determined by subtracting the end time of the preceding task from the start time of the subsequent task. Intervals between 0 and 10 seconds are marked as tightly connected; intervals exceeding 60 seconds are marked as loosely connected; and intervals between 10 and 60 seconds are marked as moderately connected. After determining the connection strength between all task segments, a correspondence is established between connection types and task segment execution order, forming a task connection relationship table. Further analysis is then conducted to determine if there are any disordered or overlapping execution behaviors between task segments. The judgment criteria are as follows: If the start time of a subsequent task is earlier than the end time of the previous task, it is considered an interleaved task, marked and investigated. After statistically analyzing the connection relationships of all task segments, the connection category, start-end time difference, and path segment number of all adjacent task segments are extracted and summarized in the time connection sequence. For example, in the robot's execution path, the end time of task segment 1 is 09:45:20, the start time of task segment 2 is 09:45:24, and the interval is 4 seconds, which is a tightly connected task segment. However, the start time of task segment 3 is 09:47:35, which has an interval of more than 2 minutes with the previous segment, and is recorded as a loosely connected segment. This information is added to the task time connection sequence for subsequent density judgment reference.

[0071] The task density judgment submodule determines the distribution range and number of task segments within each time interval based on the task time connection sequence, compares the continuity and dense distribution of adjacent task segments, identifies time intervals where tasks are frequent or scattered, and adjusts the criteria for dividing task intervals to obtain the distribution characteristics of task segments.

[0072] The start and end times of all task segments are divided along the timeline, with each segment divided into 10-minute intervals. The number of task segments within each time interval is counted, and the block number of the start and end times of each task segment is recorded. The statistical results are used to form a task density distribution table. The number of tasks in adjacent time intervals is then compared. If the increase in the number of tasks in the later interval exceeds 50% compared to the previous interval, it is identified as a task-concentrated area. If the total number of tasks in two consecutive time intervals is less than 3, it is identified as a task-sparse area. In addition, the time interval between each task segment is analyzed. If the interval between three consecutive task segments is less than 10 seconds and they are all within the same time interval, the interval is marked as a high-density segment area. Continuity is then determined in conjunction with the task connection type. If three consecutive task segments are all tightly connected or moderately connected, they are further identified as continuous dense segments. For the marked high-density or sparse blocks, the time window parameters in the original task interval division criteria are adjusted. For example, the original maximum task time width was set to 600 seconds, which is adjusted to 300 seconds in high-density areas and widened to 900 seconds in sparse areas. Through such dynamic adjustments, the system redefines the belonging relationship of each task segment on the time axis and outputs updated task segment distribution characteristics. For example, if there are 7 task segments between 09:00 and 09:10 with an average task interval of 8 seconds, this time period is identified as a high-density segment area. The task division parameter is reduced from the original 600 seconds to 300 seconds, the task boundaries are redefined, and new segment distribution information is formed.

[0073] Please see Figure 4 The node adjustment module includes:

[0074] The remaining power monitoring submodule analyzes the time series and remaining power data of the task segments based on the distribution characteristics of the task segments, calculates the rate of power change between adjacent sampling points, determines the correspondence between power change and time process in each task segment, optimizes the power data collection method, and obtains the power change rate sequence.

[0075] Extract the start and end times corresponding to each task segment to determine the position distribution of each task segment on the timeline. Then, call the remaining power collection data recorded in the battery management system to extract the power value sequence consistent with the task timestamp. Before processing the power data of the task segment, first align the power data with time at a sampling interval of 1 second to ensure that each task segment has a complete power time sequence. Then, calculate the power difference between adjacent sampling points and divide it by the time interval to obtain the power change rate. Group the power change rate values ​​by task segment number and traverse them. For segments with abnormal rate values, determine whether they belong to valid power change by setting upper and lower thresholds. Among them, the power decrease rate is less than 0.01% / s and is considered to be in the current stable zone, and greater than 0.01% / s and greater than 0.01% / s. A rate of 10% / s is identified as a rapid power consumption zone. This determination is based on historical energy consumption data of the robot during both no-load and full-load operation. For example, under conditions of a 15kg load and an average speed of 1.2m / s, if the battery level drops from 92.6% to 89.8% in 180 seconds, the average battery change rate is calculated to be 0.015% / s. This value falls within the low-speed power consumption zone, which is considered a stable power consumption segment. Then, for each task segment, the battery change rate sequence within each segment is re-collected, and a rate list is generated in chronological order. The trend of change within consecutive segments is analyzed sequentially, the trend characteristics of the speed value change curve are recorded, and the key turning points of the battery curve for each task segment and their corresponding times are marked. The battery change rate sequence is then output, collected by task segment number.

[0076] The power trend comparison submodule compares the power change rate sequence with the sampled power output data of the task segment to analyze the synchronization characteristics of power change and power change rate at each moment, using the formula:

[0077] ;

[0078] Obtain the synchronization difference index ,in, This indicates the number of sampling moments, i.e., the total number of data points collected within the task segment. Indicates the first Output power at each sampling time, Indicates the first Output power at each sampling time, Indicates the first Remaining battery power at each sampling time. Indicates the first Remaining battery power at each sampling time. This represents the time interval between two adjacent sampling times;

[0079] The synchronization difference index refers to the average absolute difference between the change in output power and the change in remaining power per unit time within a task segment. This reflects the degree of synchronization between the change in the robot's battery output power and the actual rate of energy consumption at each sampling moment. A smaller value indicates that the trends in power output change and power consumption rate are relatively consistent, meaning the dynamic relationship between energy consumption and output is relatively synchronized; if A larger value indicates a significant deviation in the trends of the two, indicating that energy output and consumption are out of sync. The indicator can be used to assess the dynamic coupling state between battery energy consumption and output load during the operation of different task segments, providing a basic quantitative basis for subsequent steps such as judging energy consumption anomalies and optimizing the selection of charging nodes.

[0080] The sampling points are set to equal intervals, with a time interval of [time value missing]. Seconds, calculate the difference between adjacent power data. Then calculate the difference between adjacent electricity data and divide by the time interval. The absolute value of the difference between the two is used to represent the deviation between the power output and the power consumption rate within each sampling interval. Finally, the average of the deviations across all sampling intervals is used to obtain the synchronization difference index. The data is standardized using the [0, 1] linear normalization method. The normalization formula is as follows:

[0081] ;

[0082] in, The original data, and These are the minimum and maximum values ​​of the dataset, with four sampling points. The original power sequence is... W, obtained after normalization The original power sequence is Ah, after normalization, we get Based on the normalized data, calculations are performed according to the formula, where... .

[0083] Calculation of the second sampling point:

[0084] ;

[0085] Calculation of the 3rd sampling point:

[0086] ;

[0087] Calculation of the 4th sampling point:

[0088] ;

[0089] The average of the sums of the three terms is:

[0090] ;

[0091] The intervals of synchronization difference are divided into the following three categories:

[0092] Interval 1: Good synchronization ( Within this range (≤0.5), the synchronization between power change and charge change rate is good, the deviation is small, and the system operates smoothly. This usually occurs during task segments where the load is stable and the battery energy consumption changes in line with the output power change.

[0093] Interval 2: Moderate synchronicity (0.5 < Within the range of ≤0.8, there is a moderate asynchrony between the rate of power change and the rate of charge change, with a significant deviation. This situation typically occurs during task periods with large load fluctuations, where the matching relationship between battery consumption and output is less affected, and the system's energy management performance is moderate.

[0094] Interval 3: Poor synchronicity ( (>0.8) Within this range, the difference between the rates of change of power and energy is significant, indicating a substantial mismatch between energy output and consumption during the task segment. This situation typically occurs under conditions of drastic load changes, sudden operating conditions, or reduced battery efficiency, requiring further optimization of task scheduling and battery management.

[0095] According to the calculation results, The deviation between the rate of power change and the rate of battery consumption change in the task segment falls within the range of poor synchronization (range 3). This indicates that the system's energy management performance is poor in this task segment, and the synchronization between power output and battery consumption is poor. Further analysis and optimization of charging scheduling and energy management strategies are needed.

[0096] The trigger node extraction submodule uses the synchronization difference index to filter out time periods that show synchronization anomalies, determines the task identifier and power change characteristics in the corresponding segment, and jointly analyzes the changes in the rate of power decrease to obtain the energy consumption dynamic control node.

[0097] The process involves filtering out periods with abnormal fluctuations from the power rate change sequence. The criteria are: if the power rate fluctuation exceeds 0.08% / s within three consecutive sampling points, and the directions of change are inconsistent (alternating between positive and negative), it is marked as a synchronous anomaly segment. After filtering, for each marked segment, the corresponding task segment identifier information is extracted, and the output power data sequence within that segment is obtained. The maximum difference within the segment is calculated for the power value change amplitude. If the power value changes by more than 15W within the same segment, it indicates a sudden change in output power. For such segments, the corresponding power rate decrease is further extracted to determine if there is a sudden jump or drop in the power rate decrease per unit time. If the difference between a power rate point and its previous point exceeds 0.05% / s, it is considered to have an abnormal fluctuation. In the case of sudden changes, task segments that meet the criteria are marked as candidate triggerable nodes. All candidate segments are cross-checked. If, in the same segment, there is a power change greater than 15W and a power rate change greater than 0.05% / s, and the task segment is in a high-density zone (task segment interval less than 20 seconds), then the segment is confirmed as a dynamic energy consumption control node. For example, during the operation of the third task segment, the power rate suddenly increases from 0.012% / s to 0.093% / s and the power output fluctuates from 35W to 58W within the time interval from 10:22:40 to 10:22:52. Since this task segment is a continuous dense scheduling segment, this period is identified as a valid trigger node, and the node occurrence time of 10:22:45 is recorded as the output result of the subsequent control node.

[0098] Please see Figure 5 The buffer discrimination module includes:

[0099] The energy consumption trajectory extraction submodule analyzes the battery energy changes and time series of each task segment based on the energy consumption dynamic control node, calculates the fluctuation range of energy consumption in continuous time periods during operation, optimizes the stability of energy consumption data, identifies the energy consumption fluctuation characteristics under the operation state, judges the influence of abnormal fluctuation signals, and obtains the energy consumption fluctuation interval sequence.

[0100] Extract the start and end times of the task segment for each control node, read the remaining battery power data within that time period, and construct an energy time series according to the sampling time interval. Perform a linear expansion of the power time series, calculate the power difference every second, and then sequentially record the energy consumption changes within consecutive time periods. Then, using each 5-second segment as a sliding window, calculate the range of power decreases within the window as the benchmark for judging energy consumption fluctuations within that window. Next, iterate through the complete timeline of each task segment, superimposing all ranges within the sliding windows by time to form an energy consumption fluctuation sequence within the task segment. Finally, sequentially determine whether the fluctuation values ​​fall into an abnormal range, setting a fluctuation amplitude of less than 0.02% as stable. The system defines segments as those with fluctuations greater than 0.1%, identifying abnormal fluctuation segments. By comparing the fluctuation type markers of three consecutive windows to determine if they are consecutively abnormal, the time range of continuous abnormal fluctuations within a task segment can be identified. For example, in task segment T005, the fluctuation amplitudes of each 5-second window between 10:14:00 and 10:14:20 are 0.13%, 0.15%, and 0.17%, respectively, all greater than 0.1%, thus being determined as abnormal continuous segments. The upper and lower boundaries of the abnormal segments are then identified to determine the start and end times of each abnormal fluctuation segment. Each abnormal segment within each task segment is marked individually. After completion, all abnormal segments in all task segments are summarized, and the energy consumption fluctuation interval sequence arranged in chronological order is output.

[0101] The state mapping analysis submodule calculates the acceleration and velocity parameters of each task segment based on the energy consumption fluctuation interval sequence, compares the synchronization relationship between energy consumption and state characteristics under each operating state, judges the synergy between energy consumption change trend and state characteristics, determines the time period with prominent synchronization, and obtains the state synchronization interval sequence.

[0102] The task segment number corresponding to each energy consumption fluctuation range is read sequentially. Then, the acceleration and velocity parameter data of the corresponding time period in that task segment are retrieved. The acceleration is calculated from the raw data of the triaxial sensor, and the velocity is converted from the encoder displacement difference result. The acceleration and velocity data are synchronously organized along the time axis. Then, the state vector at each time point is calculated, that is, the velocity value and the acceleration value are paired to form a task state sequence. Then, the fluctuation analysis of the state value of each time period in the energy consumption fluctuation sequence is performed sequentially. It is counted whether the change amplitude of acceleration in each state data segment is higher than 0.5 m / s² and whether the change value of velocity is higher than 0.8 m / s. If both conditions are met, it is marked as a state change segment. The system continues to determine whether the occurrence time of the state change segment coincides with the energy consumption fluctuation segment. If a segment with an overlap rate greater than 50% appears, the segment can be identified as a state coordination segment with strong synchronization. For example, in task segment number T008, the power fluctuation segment is from 11:20:40 to 11:21:10. During this segment, the acceleration jumps from 0.3m / s² to 1.1m / s², and the velocity increases from 1.5m / s to 2.4m / s. The duration of this state change is 28 seconds, which completely coincides with the energy consumption fluctuation segment. Therefore, this segment is identified as a state synchronization segment. All task segment segment numbers, start and end times, and state value fluctuation amplitudes that meet the synchronization criteria are recorded and output to form a state synchronization interval sequence.

[0103] The buffer segment calibration submodule, based on the state synchronization interval sequence, determines the persistence characteristics of energy consumption fluctuations, compares the coupling offset between energy consumption and motion state within each time segment, and uses the following formula:

[0104] ;

[0105] The coupling offset magnitude is obtained, and segments where the coupling offset magnitude converges continuously are identified to obtain the energy consumption buffer markers, where, Indicates the sequence number is The coupling offset magnitude of the task segment, Representing a task fragment The Middle Energy consumption rate per unit time period at each time point Representing a task fragment The Middle Acceleration parameters at each time point, Representing a task fragment The Middle Speed ​​parameters at each time point This represents the equivalent mass parameter, used to convert dynamic disturbance terms to the energy consumption level. Representing a task fragment Number of time nodes within;

[0106] The coupling offset amplitude refers to the average absolute difference between the rate of change of energy consumption per unit time and the dynamic disturbance term determined by acceleration and velocity parameters (after transformation by the equivalent mass weight of the system) within a certain task segment. The smaller the value, the more synchronized or matched the energy consumption and motion state are. The larger the value, the weaker the correspondence between the energy consumption change and the motion state, and the more significant the fluctuation or inconsistency. It can quantify the real-time response relationship between battery energy consumption and actual motion behavior. It is an important basic data for identifying energy consumption anomalies, judging the stability of task segments, or selecting charging buffer periods.

[0107] After the task segments are divided, the raw data for each time point in each segment is extracted sequentially, including the energy consumption rate. acceleration ,speed and equivalent mass parameters ,in, Data is acquired by the battery management system at a sampling frequency of 1Hz, in units of Wh / s. Recorded by the inertial measurement unit, in m / s². Calculated from the encoder, in m / s. The equivalent mass parameter for the robot is set at 48.3 kg, and the sample range is [missing value]. kg, using min-max normalization, maps each data point to The interval is calculated as follows: According to this rule:

[0108] Energy consumption rate sequence:

[0109] ;

[0110] After normalization, it becomes:

[0111] .

[0112] Acceleration:

[0113] ;

[0114] After normalization, it becomes:

[0115] .

[0116] speed:

[0117] ;

[0118] After normalization, it becomes:

[0119] .

[0120] Equivalent mass parameters In the sample interval After internal normalization, we get .

[0121] The normalized values ​​are then substituted into the formula for calculation, with each term performed as follows:

[0122] Item 1 ;

[0123] Item 2 ;

[0124] Item 3 ;

[0125] Item 4 ;

[0126] Item 5 ;

[0127] Adding all offsets together Then divide by the number of time points. The average offset amplitude is obtained as follows:

[0128] ;

[0129] Coupling offset amplitude The following preset reference range is set to determine the energy consumption and power synchronization status type of task segments:

[0130] when The section was identified as a coupling convergence zone. Within this zone, the energy consumption rate deviated little from the motion state, and the trends were basically consistent. This indicates that the energy consumption and the robot's motion load showed strong synchronicity, which is a typical energy consumption stable zone.

[0131] when It is determined to be a coupling transition section, where there is a certain degree of asynchronous fluctuation between energy consumption and state, which cannot yet be regarded as stable, but the deviation is not significant, and it is an intermediate section with adjustment potential;

[0132] when The section is identified as a coupling mismatch section, where the response relationship between energy consumption fluctuations and acceleration and speed status is weak, indicating system disturbances, sudden task loads, or disconnection of the execution mechanism. Therefore, it must be excluded from the candidate range for charging scheduling.

[0133] The obtained normalized coupling offset amplitude is Based on the above interval determination logic, this value falls into the first category, i.e. Therefore, this result indicates that the current task segment is a "coupled convergence segment", and its energy consumption change is highly consistent with the motion state, indicating that there is a characteristic that dynamic disturbances are effectively converted into energy consumption within this segment. Based on this, the segment is marked as an energy consumption buffer segment.

[0134] Please see Figure 6 The time-period filtering module includes:

[0135] The task gap identification submodule analyzes the sequential order of continuous task segments in the time series based on the energy consumption buffer mark, determines the gaps between each task segment, filters time segments with interval characteristics, and obtains a set of task interval time segments.

[0136] Extract the start and end time information of all task segments, sort them in ascending order to generate a continuous timeline sequence, then read the time boundaries between energy consumption buffers and insert them as judgment intervals into the task segment sequence. Next, calculate the interval time for each pair of adjacent task segments, i.e., subtract the end time of the previous task from the start time of the next task to obtain the actual gap time. Judge all calculation results; if the interval exceeds 15 seconds, it is marked as a valid task interval segment. The judgment criteria are the current scheduling frequency and the statistical value of the shortest response time for task scheduling in historical data, excluding scheduling task time points with delayed start, such as during the task... Task segment T003 ends at 09:32:20, and task segment T004 begins at 09:33:05. There is an actual interval of 45 seconds between the two segments, so this time period is marked as a valid task gap. Then, all time segments that meet the interval characteristics are constructed into an initial set of task interval time segments. Further, it is determined whether the interval time segments overlap or partially cover the energy consumption buffer segment. The interval segments that completely overlap are retained as high-priority gaps. If there is only partial intersection, it is determined whether the intersection length exceeds 10 seconds. If it does not reach this threshold, the segment is removed. The finally retained interval segments are uniformly recorded and output as a set of task interval time segments.

[0137] The path segment localization submodule determines the distribution of path coordinates during each interval based on the task interval time segment set, analyzes the frequency of occurrence of trajectory segments in each task segment, identifies independently distributed and non-repeating trajectory segments, and excludes path segments with overlap to obtain an exclusive segment sequence.

[0138] The robot's path coordinate data is read sequentially within each time interval. Coordinates for each path are recorded in 1-second increments, forming a trajectory sequence composed of timestamps and coordinate points. Each interval's path sequence is numbered and uniformly encoded. Next, frequency matching is performed on each interval path across all task segments, counting the number of times each coordinate point in that interval appears in the task segment's path. Coordinate points with zero occurrences are classified as independent path points. If independent path points account for more than 90% of the entire path segment, it is considered an exclusive path segment; otherwise, overlap is calculated. If the overlap is high... If 20% or more are found to be duplicate segments, the path segment is excluded and no longer retained as an exclusive path segment. For example, in gap segment S005, the total path length is 36 meters, and 32 meters of the corresponding coordinate point sequence do not appear in any task path, with an independent percentage of 88.9%, which does not meet the 90% judgment criterion. Therefore, this path segment is removed. In another segment S006, the path length is 24 meters, and 23.5 meters of the coordinate points do not appear in the task segment, with an independent percentage of 97.9%. Therefore, it is retained as a valid exclusive path segment. Finally, all retained path segments will be uniformly classified and marked as exclusive segment sequences.

[0139] The available interval filtering submodule compares the positional relationship between each candidate charging segment and the time interval of the scheduling sequence task based on the exclusive segment sequence, analyzes the overlap between the candidate segments and the task time interval, identifies the charging available segments that do not overlap, optimizes the scheduling interval rules, and obtains the charging scheduling available interval.

[0140] First, the start and end times of each exclusive path are read and cross-checked with the time intervals of each task segment in the scheduling sequence. Each exclusive path's time interval is considered a candidate charging segment. The process involves comparing the candidate segments with the task time intervals to determine if there is any overlap. During this process, the start and end times of each pair of time intervals are intersected. If the intersection is empty (meaning no time points overlap), the segment is considered valid and usable. If overlap exists, the overlap duration is checked to see if it exceeds 30% of the total duration of the candidate segments. If it does, the segment is excluded; otherwise, the non-overlapping portion is retained as a usable segment. For example, if an exclusive segment P007 has a time of 14:2... From 2:00 to 14:25:30, a total of 210 seconds, there is an 80-second overlap with the task segment T009 time period from 14:23:10 to 14:24:30, with an overlap ratio of 38%, which exceeds the threshold. Therefore, P007 is excluded. Another path, P008, is between 15:02:00 and 15:04:00, which only overlaps with the task segment for 30 seconds, accounting for 25%. The non-overlapping segments of 14:22:00 to 14:23:10 and 14:24:30 to 14:25:30 are retained as two valid segments and added to the available interval list. All exclusive path time periods that do not intersect with the task time or whose intersection ratio does not exceed the threshold are output as available intervals for charging scheduling.

[0141] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A battery charging time planning system based on cycling behavior analysis, characterized in that, The system comprises: An inertia recognition module, based on robot operation information, analyzes historical paths, compares trajectory differences combined with load changes, recognizes motion mutations and path continuous segments, and determines path dynamic feature nodes according to navigation data and node changes; The inertia recognition module comprises: A data stream receiving sub-module, based on robot operation information, analyzes attitude information and angular velocity information collected by an inertia measurement device, combines acceleration and speed information recorded by an acceleration sensor, compares the change trends of various data in the same time period, identifies time segments associated with path direction changes, and obtains inertia response segments; A path load comparison sub-module, based on the inertia response segments, compares trajectory data under various load conditions, analyzes the differences in path direction changes in each load interval, identifies segments with prominent angular velocity changes, determines trajectory deviation characteristics in each interval, and obtains path deviation patterns; A node evolution extraction sub-module, based on the path deviation patterns, analyzes motion mutation segments and continuous trajectory segments, determines the position changes of path nodes in each time period, compares the time sequence distribution characteristics between nodes, identifies representative dynamic migration trajectories, and obtains path dynamic feature nodes; A density recognition module, based on the path dynamic feature nodes, determines the start and end nodes of each task segment, analyzes the time connection characteristics between task segments, compares the time span differences between segments, identifies task-intensive and task-sparse intervals, and obtains task segment distribution characteristics; A node adjustment module, based on the task segment distribution characteristics, analyzes the changes of remaining power and output power in the task segment, determines the synchronization relationship between rhythm changes and energy consumption, adjusts the charging start determination basis, and obtains energy consumption dynamic control nodes; The rhythm changes refer to the change rhythm of task intensity or behavior patterns during task execution; A buffer discrimination module, based on the energy consumption dynamic control nodes, analyzes the energy consumption change characteristics of the task segment, determines the relationship between energy consumption and operating state, compares energy consumption distribution, identifies energy consumption buffer segments, and obtains energy consumption buffer interval markers; A time period screening module, based on the energy consumption buffer interval markers, analyzes the distribution of path repetition segments, compares the coverage of available charging segments in the scheduling sequence, identifies non-overlapping charging intervals, and obtains charging scheduling available intervals.

2. The battery charging time planning system based on cycling behavior analysis of claim 1, wherein, The path dynamic feature nodes include motion response attributes, node distribution states, and trajectory change types, the task segment distribution characteristics include distribution interval attributes, segment coherence characteristics, and task intensity, the energy consumption dynamic control nodes include control trigger times, node corresponding energy consumption levels, and control reference parameters, the energy consumption buffer interval markers include buffer identification segments, energy consumption change intervals, and task state labels, and the charging scheduling available intervals include charging allocation time periods, task interval blocks, and available scheduling labels. 3.The battery charging time planning system based on cycling behavior analysis of claim 1, wherein, The density recognition module comprises: A start and end node recognition sub-module, based on the path dynamic feature nodes, determines the arrangement of adjacent nodes in the navigation sequence and the task corresponding state, screens node combinations with task boundary characteristics in the navigation sequence, identifies the start and end nodes of each task segment, and obtains the task segment start and end nodes; The time connection structure analysis submodule analyzes the time connection characteristics between the continuous task fragments based on the task fragment start and end nodes, judges the execution sequence and connection degree inside and outside the task fragments, identifies the connection fragments with different time spans, and obtains a task time connection sequence; The task density judgment submodule judges the distribution range and quantity of the task fragments in each time interval based on the task time connection sequence, compares the continuity and dense distribution of adjacent task fragments, identifies the time sections with frequent or scattered tasks, and adjusts the task interval division basis to obtain task fragment distribution characteristics. 4.The battery charging time planning system based on cycling behavior analysis of claim 1, wherein, The node adjustment module comprises: The residual power monitoring submodule analyzes the time sequence of the task fragments and the residual power collection data based on the task fragment distribution characteristics, calculates the power change rate between adjacent sampling points, judges the corresponding relationship between the power change and the time process in each task fragment, optimizes the power data collection mode, and obtains a power change rate sequence; The power trend comparison submodule compares the sampling data of the task fragment output power based on the power change rate sequence, analyzes the synchronization characteristics of the power change and the power change rate at each time, and obtains a synchronization difference index; The trigger node extraction submodule screens the time sections that exhibit synchronization abnormalities based on the synchronization difference index, judges the task identification and power change characteristics in the corresponding sections, and jointly analyzes the change of the power decline rate to obtain an energy consumption dynamic regulation node. 5.The battery charging time planning system based on cycling behavior analysis of claim 1, wherein, The buffer discrimination module comprises: The energy consumption trajectory extraction submodule analyzes the battery energy change and time sequence of each task fragment based on the energy consumption dynamic regulation node, calculates the fluctuation range of the energy consumption in the continuous time section during the operation, optimizes the stability of the energy consumption data, identifies the energy consumption fluctuation characteristics in the running state, judges the influence of abnormal fluctuation signals, and obtains an energy consumption fluctuation interval sequence; The state mapping analysis submodule calculates the acceleration and speed parameters of each task fragment based on the energy consumption fluctuation interval sequence, compares the synchronization relationship between the energy consumption and the state characteristics in each running state, judges the cooperativity of the energy consumption change trend and the state characteristics, determines the time sections with prominent synchronization, and obtains a state synchronization interval sequence; The buffer section calibration submodule judges the duration characteristics of the energy consumption fluctuation based on the state synchronization interval sequence, compares the coupling offset of the energy consumption and the motion state in each time section, and uses the formula: ； The coupling offset amplitude is obtained, and a section with continuous convergence of the coupling offset amplitude is identified to obtain an energy consumption buffer interval marker, wherein, represents the coupling offset amplitude of a task section with a sequence number of represents a unit time period energy consumption rate of a time node with a number of in the task section represents an acceleration parameter of a time node with a number of represents a speed parameter of a time node with a number of in the task section represents an equivalent mass parameter, represents a number of time nodes in the task section .​​​​​ 6.The battery charging time planning system based on cycling behavior analysis of claim 1, wherein, The time section screening module comprises: The task gap identification submodule analyzes the sequence of the continuous task fragments in the time sequence based on the energy consumption buffer interval marker, judges the gaps between the task fragments, screens the time sections with interval characteristics, and obtains a task interval time section set; The path fragment positioning submodule judges the path coordinate distribution in each interval based on the task interval time section set, analyzes the occurrence frequency of the trajectory fragments in each task fragment, identifies the trajectory sections that are independently distributed and not repeated, and excludes the path sections with overlapping degrees to obtain an exclusive section sequence; The available interval screening sub-module compares the position relationship between each candidate charging segment and the task time interval of the scheduling sequence based on the exclusive segment sequence, analyzes the overlap between the candidate charging segment and the task time interval, identifies the charging available interval without intersection, optimizes the scheduling interval rule, and obtains the charging scheduling available interval. 7.The battery charging time planning system based on cycling behavior analysis of claim 1, wherein, The historical path refers to the path trajectory walked by the robot during work or test, including a coordinate point sequence and time sequence, each task segment refers to a continuous motion interval divided on the path and having independent execution meaning, and the start and end nodes refer to the positions on the path or task sequence at the beginning and end of each task segment.

Images