A method and system for lawn management based on internet of things

Abstract

The application relates to the technical field of Internet of Things sensing management, in particular to a lawn management method and system based on Internet of Things sensing. The method comprises the following steps: acquiring lawn structure data and Internet of Things sensing terminal data, constructing a lawn structure topology graph and a lawn weed candidate evolution context, and generating context terminal association data; generating a context observation record time sequence and calculating evolution index data, generating a cross-sensing task of the lawn weed candidate evolution context according to the evolution index data, identifying observed lawn weeds by executing the cross-sensing task through the Internet of Things sensing terminal; selecting key evolution context segments that meet context segment screening rules, determining task configuration parameters of the key evolution context segments, and generating a context management task list; and generating terminal execution instruction data for each key evolution context segment matched with a target terminal. The application realizes a precise lawn management based on Internet of Things sensing, which is lawn weed fine and region clear.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) sensing and management technology, specifically to an IoT-based lawn management method and system. Background Technology

[0002] In urban parks, sports fields, residential green spaces, and median strips, lawns are a common form of greening and require long-term weed monitoring and control. Current technologies primarily employ periodic manual inspections combined with zonal spraying or mechanical cutting for lawn weed control. Some scenarios have incorporated IoT devices such as video surveillance terminals, mobile inspection terminals, or spraying robots. Traditional management methods typically divide lawns into regular grids or management zones, arranging inspection routes and spraying operations on a grid-by-grid basis, and configuring observation tasks on a single-point terminal basis. In terms of time organization, routine operations are performed according to a unified maintenance cycle. There is a lack of mechanisms for modeling the correlation between weed growth and spread behavior and structural elements such as internal irrigation pipelines, drainage paths, walking paths, and hard boundaries within the lawn. Furthermore, there is a lack of pathways to continuously express the field-of-view coverage relationship of IoT terminals and weed monitoring results. In actual lawn maintenance and management, there are generally sprinkler lines, drainage lines, pedestrian paths, and boundaries between lawns and hard paving within the lawn. This presents a problem where weeds recur and spread along these internal structural lines, but existing management models cannot precisely identify and configure control tasks according to these path segments. Specifically: Firstly, weed observation results collected by multi-source IoT sensing terminals are mainly recorded at scattered time points and discrete spatial locations, lacking a continuous evolution path organized along the lawn's structural lines. This makes it difficult to identify key weed sections that have formed long-term, recurring, and directional spread. Secondly, IoT operational information such as sprinkler plans, maintenance operation plans, and the operational status of control terminals are isolated from weed observation results. There is a lack of a mechanism to uniformly map observation capabilities, operational constraints, and weed evolution characteristics into an executable control task list based on path segments. This results in some areas near structural lines experiencing long-term excessive or repeated spraying yet still experiencing recurring weed growth, while other areas near structural lines lack targeted observation and control for extended periods. Summary of the Invention

[0003] The purpose of this invention is to provide an IoT-sensing lawn management method and system to solve the problem mentioned in the background art that weeds recur and spread along the internal structural lines, but existing management models cannot accurately identify and configure control tasks according to path segments.

[0004] To achieve the above objectives, the technical solution of the present invention is: an IoT-sensing lawn management method, comprising: S1. Acquire lawn structure data and IoT sensing terminal data, construct a lawn structure topology map based on the lawn structure data, extract structural line elements from the lawn structure topology map to construct a candidate evolutionary network of lawn weeds, calculate the observable relationship between the candidate evolutionary network of lawn weeds and the IoT sensing terminal based on the IoT sensing terminal configuration data, and generate network terminal association data. Among them, the candidate evolutionary network of lawn weeds is a linear topological object that describes the relationship of the growth and spread paths of weeds distributed along the lawn structure line within the lawn area, and is used to define the observation range and operation range of IoT sensing terminals. S2. Based on the network terminal association data, determine the set of responsible terminals for the candidate evolutionary network of lawn weeds, generate the network observation record time series and calculate the evolution index data, generate the cross-sensing task of the candidate evolutionary network of lawn weeds according to the evolution index data, and perform the cross-sensing task through the Internet of Things sensing terminal to identify the observed lawn weeds and update the network observation record time series. Among them, the network observation record time series is a data sequence that records the IoT sensing terminal identifier, observation location and weed observation results of candidate evolutionary networks of lawn weeds at each observation time in chronological order. S3. Calculate evolutionary feature data based on the time series of the network observation records, select key evolutionary network segments that meet the network segment screening rules, read the IoT operation status data associated with the spatial location of the key evolutionary network segments, determine the task configuration parameters of the key evolutionary network segments, and generate a network management task list. Among them, the key evolutionary network segment is the network interval with continuous weed growth and spread along the network direction identified on the candidate evolutionary network of turfgrass weeds. S4. Match target terminals for each key evolutionary segment according to the context management task list, and generate terminal execution instruction data corresponding to the key evolutionary segment.

[0005] Preferably, in S1, the lawn structure topology map is a structural map composed of lawn area boundaries, internal facility locations, and walking path nodes as topological nodes and the lines connecting the nodes as topological edges. It is used to characterize the connection and spatial relationship between lawn area boundaries, sprinkler irrigation facilities, drainage facilities, and walking paths. The structural line elements are the basic linear data for constructing the candidate evolutionary network of lawn weeds. The candidate evolutionary network of lawn weeds is also used as an index unit for the time series of organizational network observation records, and as an association identifier between the set of responsible terminals, key evolutionary network segments, and the network management task list.

[0006] Preferably, in S1, the observable relationship between the candidate evolutionary network of lawn weeds and the IoT sensing terminal is determined based on the installation location, field of view, installation height, and effective observation distance of the IoT sensing terminal. This relationship defines the observable interval of the candidate evolutionary network of lawn weeds within the field of view of the IoT sensing terminal and the corresponding observation conditions. This is used to limit the observable segment of the IoT sensing terminal on each candidate evolutionary network of lawn weeds. The network terminal association data is generated based on the observable relationship and records the identifier of the candidate evolutionary network of lawn weeds, the identifier of the IoT sensing terminal, the start and end positions of the observable interval of the IoT sensing terminal on the corresponding candidate evolutionary network of lawn weeds, and the observation time constraints. This data serves as the basis for generating the cross-sensing task.

[0007] Preferably, in step S2, the responsible terminal set is a subset of terminals selected from IoT sensing terminals that have observable intervals with the candidate evolutionary network of lawn weeds based on the network terminal association data. It includes a main observation terminal covering the target observation interval of the candidate evolutionary network of lawn weeds and multiple collaborative observation terminals covering overlapping intervals. These terminals are used to perform observation tasks on the candidate evolutionary network of lawn weeds and provide terminal resources for cross-sensing tasks. The selection criteria for the responsible terminal set include the length of the observable interval of the IoT sensing terminal on the corresponding candidate evolutionary network of lawn weeds, the coverage ratio of the observable interval, and the terminal observation reliability index. The responsible terminal set is formed by comprehensively sorting the observable interval and the observation reliability index and selecting IoT sensing terminals that meet the preset coverage conditions.

[0008] Preferably, in S2, the network observation record time series is a data sequence that records the IoT sensing terminal identifiers, observation times, observation location parameters, and weed observation results observed during the execution of the responsible terminal set and cross-sensing tasks, using the lawn weed candidate evolution network as an index and in chronological order. This sequence reflects the weed distribution status of the lawn weed candidate evolution network at different observation times and provides input for the calculation of evolution index data and evolution feature data. Specifically, the data structure of the network observation record time series is a list of observation records sorted by observation time, with the lawn weed candidate evolution network identifier as the primary key. Each observation record includes at least an observation timestamp, IoT sensing terminal identifier, network location parameters, and weed observation result fields. The generation method of the network observation record time series is as follows: during the execution of the observation tasks and cross-sensing tasks issued based on the responsible terminal set, the observation data returned by the IoT sensing terminals is processed by weed identification and then sequentially traced... The observation records are added to the list of candidate evolutionary trajectories for the corresponding lawn weeds, and the observation records are sorted and updated according to the observation timestamp. The evolution index data is a set of statistical indicators calculated based on the time series of the trajectories observation records within a preset statistical time window. The evolution index data includes the observation frequency of candidate evolutionary trajectories for lawn weeds, the consistency index of observation results, and the observation time window index, which are used to guide the generation of cross-sensing tasks and the adjustment of observation tasks. The calculation method of the evolution index data is as follows: the number of observation records in the time series of trajectories observation records within the preset time window is counted to obtain the observation frequency index; the consistency ratio of weed observation results of different IoT sensing terminals at the same or adjacent trajectories within the same time window is counted to obtain the observation consistency index; the time interval between the current time and the timestamp of the most recent observation record is calculated to obtain the observation time window index; and the observation frequency index, the observation consistency index, and the observation time window index are combined to form the evolution index data.

[0009] Preferably, in S2, the cross-sensing task is an observation task generated based on evolutionary index data for a target observation interval where the consistency index of the observation results in the candidate evolutionary network of lawn weeds is lower than a preset threshold and the observation time window index is higher than a preset threshold. The cross-sensing task includes lawn weed candidate evolutionary network identifiers, target observation interval parameters, participating IoT sensing terminal identifiers, and cross-observation time window parameters, which are used for multiple IoT sensing terminals to perform supplementary observations within the same target observation interval and within the cross-observation time window.

[0010] Preferably, in step S3, the evolutionary feature data is a set of feature parameters calculated based on the time series of the network observation records within multiple observation periods. It is used to characterize the first appearance location of weeds on the candidate evolutionary network of lawn weeds, the spread trend along the network direction, and the recurrence at local network locations. It is also used to distinguish between stable weed growth and spread sections and high-incidence and easily spread weed sections on the candidate evolutionary network of lawn weeds. The data structure of the evolutionary feature data is specifically a set of feature records that use the lawn weed candidate evolutionary network identifier as an index and record the first appearance location parameter, spread direction parameter, and recurrence frequency parameter within each preset statistical time window.

[0011] Preferably, in S3, the vein segment screening rule is as follows: on a given candidate evolutionary vein of lawn weeds, when the first occurrence position parameter within the vein position interval falls within a preset length range, and the displacement direction of the weed observation position characterized by the propagation direction parameter remains consistent along the vein direction within multiple consecutive statistical time windows, and the recurrence number parameter exceeds a preset recurrence threshold, the vein position interval is determined as a key evolutionary vein segment; if multiple adjacent position intervals satisfy the above conditions and the distance between them is lower than a preset distance threshold, then the adjacent position intervals are merged to form a key evolutionary vein segment with a larger coverage area.

[0012] Preferably, in step S3, the task configuration parameters are a set of parameters determined based on the evolutionary feature data corresponding to the key evolutionary segment and the IoT operation status data associated with the spatial location of the key evolutionary segment. These parameters include task type parameters, task execution time parameters, task priority parameters, and resource constraint parameters for the key evolutionary segment. The network management task list is a set of task records indexed by the key evolutionary segment identifier. Each task record includes the key evolutionary segment identifier, the network location parameters of the key evolutionary segment, the corresponding task configuration parameters, and the target management terminal type information.

[0013] On the other hand, the present invention provides an IoT-sensing lawn management system, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned IoT-sensing lawn management method.

[0014] Compared with the prior art, the above-mentioned technical solution of the present invention has the following beneficial technical effects: 1. In this invention, a candidate evolutionary network of lawn weeds is constructed based on the lawn structure topology map, and the observation record time series of IoT sensing terminals are organized on the evolutionary network. This can transform the weed observation data, which was originally managed in a grid or isolated terminal, into a weed growth and spread path that is continuously expressed along structural lines such as sprinkler pipelines, drainage directions, and pedestrian paths, thereby reducing blind spots and areas of repeated observation. 2. In this invention, evolutionary feature data is calculated based on the time series of network observation records. Key evolutionary network segments are selected and combined with sprinkler irrigation plans, maintenance operation plans, and management terminal operation status to generate a network management task list. This enables the configuration of management methods, execution time, priority, and other parameters by path segment as a unit. The management tasks are precisely bound to key evolutionary network segments with recurrence and spread risks, avoiding large-area uniform spraying or omission of key segments. This helps to reduce management operation input while ensuring the controllability of lawn condition. Attached Figure Description

[0015] Figure 1 This is a flowchart of one embodiment of the present invention. Detailed Implementation

[0016] Example 1, as Figure 1 As shown, the specific implementation steps of the IoT-based lawn management method proposed in this invention are as follows: S1. Acquire lawn structure data and IoT sensing terminal data, construct a lawn structure topology map based on the lawn structure data, extract structural line elements from the lawn structure topology map to construct a candidate evolutionary network of lawn weeds, calculate the observable relationship between the candidate evolutionary network of lawn weeds and the IoT sensing terminal based on the IoT sensing terminal configuration data, and generate network terminal association data. Among them, the candidate evolutionary network of lawn weeds is a linear topological object that describes the relationship of the growth and spread paths of weeds distributed along the lawn structure line within the lawn area, and is used to define the observation range and operation range of IoT sensing terminals. S2. Based on the network terminal association data, determine the set of responsible terminals for the candidate evolutionary network of lawn weeds, generate the network observation record time series and calculate the evolution index data, generate the cross-sensing task of the candidate evolutionary network of lawn weeds according to the evolution index data, and perform the cross-sensing task through the Internet of Things sensing terminal to identify the observed lawn weeds and update the network observation record time series. Among them, the network observation record time series is a data sequence that records the IoT sensing terminal identifier, observation location and weed observation results of candidate evolutionary networks of lawn weeds at each observation time in chronological order. S3. Calculate evolutionary feature data based on the time series of the network observation records, select key evolutionary network segments that meet the network segment screening rules, read the IoT operation status data associated with the spatial location of the key evolutionary network segments, determine the task configuration parameters of the key evolutionary network segments, and generate a network management task list. Among them, the key evolutionary network segment is the network interval with continuous weed growth and spread along the network direction identified on the candidate evolutionary network of turfgrass weeds. S4. Match target terminals for each key evolutionary segment according to the context management task list, and generate terminal execution instruction data corresponding to the key evolutionary segment.

[0017] In this embodiment S1, the lawn structure data includes spatial location and attribute information describing the lawn area boundary, internal sprinkler facilities, drainage facilities, light poles, manhole covers and other ancillary facilities, as well as walking paths. The lawn structure data can be exported from landscape design drawings, BIM models, GIS systems or on-site survey data, expressed using a unified plane coordinate system, with vector surface elements representing the lawn area boundary, line elements representing sprinkler and drainage pipelines, line or surface elements representing walking paths and hard paved areas, and point elements representing facility nodes such as sprinklers, valves, and inspection wells. Attribute fields record information such as facility type, pipeline type, and ground material type, providing a unified data source and coordinate basis for constructing the lawn structure topology map.

[0018] In this embodiment S1, the IoT sensing terminal data includes configuration parameters such as device identifier, installation location coordinates, installation height, orientation angle, field of view, effective observation distance, working time parameters, and terminal type for each sensing terminal. The device identifier is used to uniquely identify the sensing terminal in the system. The installation location coordinates adopt a plane coordinate system consistent with the lawn structure data. The installation height and orientation angle are used to determine the projection shape of the field of view on the ground. The field of view and effective observation distance are used to limit the observation range of the terminal on the ground. The working time parameters are used to describe the time interval during which the terminal is in an available state. The terminal type is used to distinguish different sensing forms such as fixed camera terminals, mobile inspection terminals, or vehicle-mounted terminals, thereby supporting the unified calculation and management of the observable relationship in the future.

[0019] In this embodiment S1, the lawn structure topology map is a structural map composed of lawn area boundaries, internal facility locations, and walking path nodes as topological nodes, and the lines connecting the nodes as topological edges. It is used to characterize the connection relationships and spatial relationships between lawn area boundaries, sprinkler systems, drainage systems, and walking paths. The structural line elements are sprinkler pipeline routing lines, drainage runoff lines, walking path centerlines, and lawn-hard paving boundary lines extracted from the lawn structure topology map, used as basic linear data for the candidate evolutionary network of lawn weeds. The structural line elements include sprinkler pipeline routing lines, drainage runoff lines, walking path centerlines, and grass... The boundary line between the lawn and hard paving; the candidate evolutionary network of lawn weeds is also used as an index unit for the time series of organizational network observation records, and as an association identifier between the set of responsible terminals, key evolutionary network segments and the network management task list; the construction method of the candidate evolutionary network of lawn weeds is as follows: based on the lawn structure topology map, the structural line elements are classified, the structural line elements adjacent to sprinkler irrigation facilities, drainage facilities and walking paths are selected, the selected structural line elements are spliced ​​together according to the topological connectivity relationship to form connected lines, and a unique network identifier is assigned to each connected line. The connected lines with network identifiers are used as candidate evolutionary networks of lawn weeds.

[0020] In this embodiment S1, the process of constructing a lawn structure topology map based on lawn structure data is as follows: First, the inflection points of the lawn area boundary polyline, the location points of sprinkler facilities, the location points of drainage facilities, and the inflection points of walking paths are taken as a set of topology nodes. The sprinkler lines, drainage lines, the centerline of walking paths, and the boundary lines between the lawn and hard paved areas are taken as a set of topology edges. When constructing topology edges, the starting node, ending node, and geometric shape information of each edge are maintained, and the category to which the edge belongs is marked according to the attribute field of the edge as sprinkler line, drainage line, walking path, or boundary line. Through the connection relationship between topology nodes and topology edges, a complete lawn structure topology map is formed, so that the connectivity and spatial distribution relationship between the boundaries, internal facilities, and walking paths within the lawn area can be queried and calculated through the topology map.

[0021] In this embodiment S1, the method for extracting structural line elements is as follows: In the lawn structure topology map, the topological edges of the pipeline type of sprinkler irrigation pipeline are aggregated according to the connected components to form sprinkler irrigation pipeline routing lines; the topological edges corresponding to the drainage pipeline or surface drainage direction are aggregated according to the water flow direction to form drainage confluence lines; the boundary lines of the walking path are centerlined to form the center lines of the walking path; and the intersection line between the boundary of the lawn area and the boundary of the hard paving area is taken as the boundary line between the lawn and the hard paving. The above structural line elements are classified and stored according to type, and their node sequence and geometric shape in the topology remain unchanged, so that the structural line elements can reflect the linear structure related to water migration and personnel passage in the lawn area, and can also serve as the basic linear data set when constructing the candidate evolutionary context of lawn weeds.

[0022] In this embodiment S1, the observable relationship between the candidate evolutionary network of lawn weeds and the IoT sensing terminal is determined based on the installation location, field of view, installation height, and effective observation distance of the IoT sensing terminal. This relationship defines the observable interval of the candidate evolutionary network of lawn weeds within the field of view of the IoT sensing terminal and the corresponding observation conditions. This relationship is used to limit the observable segment of the IoT sensing terminal on each candidate evolutionary network of lawn weeds. The network terminal association data is generated based on the observable relationship and records the identifier of the candidate evolutionary network of lawn weeds, the identifier of the IoT sensing terminal, the start and end positions of the observable interval of the IoT sensing terminal on the corresponding candidate evolutionary network of lawn weeds, and the observation time constraints. This data serves as the basis for generating the cross-sensing task.

[0023] In this embodiment S1, the method for constructing the candidate evolutionary network of lawn weeds is as follows: In the set of structural line elements, the irrigation pipeline routing line, drainage runoff line, walking path center line, and the boundary line between the lawn and hard paving that intersect or are adjacent to the interior of the lawn area are selected. According to the topological connectivity, structural line elements that belong to the same connected component and have the same direction or can be connected in series are spliced ​​according to the node sequence to form a connected polyline that extends continuously along the lawn structural line. A unique network identifier is assigned to each connected polyline, and the node sequence, length parameter, and spatial coordinates of the polyline are stored in the network data structure. The constructed candidate evolutionary network of lawn weeds inherits the spatial geometric characteristics and directional information of the structural line elements and serves as an index unit for the time series of organizational network observation records. It is used for linear referencing and positioning of spatial objects in the process of lawn weed observation and management configuration.

[0024] In this embodiment S1, the calculation method for the network terminal association data is as follows: First, based on the IoT sensing terminal data, a field-of-view model of each sensing terminal is established in a unified coordinate system. The terminal installation position is used as the base point for the field-of-view projection, and a fan-shaped or polygonal area of ​​the field of view is constructed on the ground plane with the field of view angle and effective observation distance. Then, the geometric intersection operation is performed between the broken line of each candidate evolutionary line of lawn and weeds and the field of view area of ​​the corresponding terminal to determine the observable interval of the line within the field of view area of ​​the terminal. The start and end positions of the observable interval on the line are recorded in the form of node sequence index or distance parameters along the line. The corresponding observation time constraints are determined by combining the terminal's working time parameters and environmental limitation parameters. Finally, the line identifier, sensing terminal identifier, observable interval start and end parameters, and observation time constraints are combined to form a network terminal association record. Multiple association records constitute a network terminal association data set, which is used to query the observable segments and available observation times of each candidate evolutionary line of lawn and weeds under different sensing terminals in subsequent steps, thereby supporting the selection of the responsible terminal set and the arrangement of sensing tasks.

[0025] In this embodiment S1, the candidate evolutionary network of lawn weeds and the associated data of network terminals are associated in the data structure through network identifiers. Each candidate evolutionary network of lawn weeds corresponds to one or more sets of network terminal associated records, and each network terminal associated record corresponds to an IoT sensing terminal and its observable range and observation time constraints. Through this structural design, while maintaining the stability of the geometric and topological information of the candidate evolutionary network of lawn weeds, the observable relationship can be independently maintained and updated according to the configuration of different sensing terminals. At the same time, it supports traversing the candidate evolutionary network of lawn weeds one by one and reverse querying from the perspective of the terminal.

[0026] In this embodiment S2, the responsible terminal set is a subset of terminals selected from IoT sensing terminals that have observable intervals with the candidate evolutionary network of lawn weeds based on the network terminal association data. It includes a main observation terminal covering the target observation interval of the candidate evolutionary network of lawn weeds and multiple collaborative observation terminals covering overlapping intervals. These terminals are used to perform observation tasks on the candidate evolutionary network of lawn weeds and provide terminal resources for cross-sensing tasks. The selection criteria for the responsible terminal set include the length of the observable interval of the IoT sensing terminal on the corresponding candidate evolutionary network of lawn weeds, the coverage ratio of the observable interval, and the terminal observation reliability index. The responsible terminal set is formed by comprehensively sorting the observable interval and the observation reliability index and selecting IoT sensing terminals that meet the preset coverage conditions.

[0027] In this embodiment S2, for each candidate evolutionary lineage of lawn weeds, a set of IoT sensing terminals with observable intervals related to the lineage is obtained based on the lineage terminal association data. For each terminal, the observable interval length, coverage ratio, and terminal observation reliability index on the target lineage are calculated. The terminal observation reliability index is obtained by statistically analyzing the terminal observation success rate, the degree of consistency with multi-terminal results, and the terminal fault reporting status in historical observation records. Based on this, the observable interval coverage ratio and the terminal observation reliability index are weighted and sorted. Terminals with longer coverage intervals and higher observation reliability are selected as primary observation terminals. Then, a group of terminals with overlapping observable intervals with the primary observation terminals and different installation locations and terminal types are selected from the remaining terminals as collaborative observation terminals. The primary observation terminals and collaborative observation terminals together constitute the responsible terminal set for the candidate evolutionary lineage of lawn weeds. During the maintenance of the responsible terminal set, the system periodically updates the terminal observation reliability index of each terminal according to the latest observation records, and removes the corresponding terminal from the responsible terminal set when the index drops below a preset threshold.

[0028] In this embodiment S2, for each candidate evolutionary line of lawn weeds, an observation task is generated based on its responsible terminal set and sent to the responsible terminals to perform periodic or triggered observations. The IoT sensing terminal collects images or other sensing data covering the target observation area according to the task requirements, and sends back the observation message containing the terminal identifier, observation time, original information of the observation location, and observation data content to the management platform. The management platform performs weed identification processing on the returned observation data to obtain weed observation result fields such as whether weeds exist, the degree of weed coverage, or the weed type number. The terminal identifier, observation time, converted into lineage position parameters along the candidate evolutionary line of lawn weeds, and weed observation result fields are assembled into an observation record and appended to the lineage observation record time series of the corresponding candidate evolutionary line of lawn weeds. When multiple terminals observe the same lineage location within a similar time period, the system writes each observation record into the time series and rearranges the time series according to the observation timestamp to support subsequent traversal analysis of the lineage observation records based on time order. At the same time, historical records with a time series length exceeding a preset upper limit are archived and stored.

[0029] In this embodiment S2, the selection of the responsible terminal set is based on a comprehensive evaluation of the observable interval and the terminal observation reliability index. A comprehensive score is calculated for each IoT sensing terminal with an observable interval in each candidate evolutionary path of lawn weeds. The comprehensive score is obtained by weighting the observable interval coverage ratio of the terminal in the target observation interval with the terminal observation reliability index according to a preset weight. The coverage ratio characterizes the spatial coverage capability of the terminal in the target observation interval, and the terminal observation reliability index characterizes the stability and reliability of the terminal in historical observations. Candidate terminals are sorted from high to low according to their comprehensive scores, and the terminals with the highest scores are selected to be added to the responsible terminal set. Each time a terminal is added, the total coverage ratio of the responsible terminal set in the target observation interval and the length of the maximum continuous uncovered segment within the target observation interval are recalculated. The selection stops when the total coverage ratio is not lower than the coverage ratio threshold and the length of the maximum continuous uncovered segment does not exceed the maximum gap threshold. The current responsible terminal set is considered as a terminal set that meets the preset coverage conditions. The coverage ratio threshold is used to stipulate that most of the length of the target observation interval should be covered by the observable interval of the responsible terminal set. The maximum gap threshold is used to limit the length of the continuous segment in the target observation interval that is not observed by any responsible terminal to not exceed the preset value, so as to reduce large-scale blind spots that cannot be observed for a long time.

[0030] In this embodiment S2, the network observation record time series is a data sequence that records the IoT sensing terminal identifiers, observation times, observation location parameters, and weed observation results observed during the execution of the responsible terminal set and cross-sensing tasks, using the lawn weed candidate evolution network as an index and in chronological order. This data reflects the weed distribution status of the lawn weed candidate evolution network at different observation times and provides input for the calculation of evolution index data and evolution feature data. Specifically, the data structure of the network observation record time series is a list of observation records sorted by observation time, with the lawn weed candidate evolution network identifier as the primary key. Each observation record includes at least an observation timestamp, IoT sensing terminal identifier, network location parameters, and weed observation result fields. The generation method of the network observation record time series is as follows: during the execution of the observation tasks and cross-sensing tasks issued based on the responsible terminal set, the observation data returned by the IoT sensing terminals is processed by weed identification and then appended sequentially. The observation records are written into the list of candidate evolutionary trajectories for the corresponding lawn weeds, and the observation records are sorted and updated according to the observation timestamp. The evolution index data is a set of statistical indicators calculated based on the time series of the trajectories observation records within a preset statistical time window. The evolution index data includes the observation frequency of candidate evolutionary trajectories for lawn weeds, the consistency index of observation results, and the observation time window index, which are used to guide the generation of cross-sensing tasks and the adjustment of observation tasks. The calculation method of the evolution index data is as follows: the number of observation records in the time series of trajectories observation records within the preset time window is counted to obtain the observation frequency index; the consistency ratio of weed observation results of different IoT sensing terminals at the same or adjacent trajectories within the same time window is counted to obtain the observation consistency index; the time interval between the current time and the timestamp of the most recent observation record is calculated to obtain the observation time window index; and the observation frequency index, the observation consistency index, and the observation time window index are combined to form the evolution index data.

[0031] In this embodiment S2, data describing the evolution of the observation state is extracted from the time series of the network observation records. A preset statistical time window is set for each candidate evolutionary network of lawn weeds. The length of the statistical time window is configured to be several days or several weeks according to the growth and inspection cycle of lawn weeds. Within each statistical time window, the network observation record time series is statistically analyzed, and the number of observation records is calculated to obtain the observation frequency index. The consistency ratio of weed observation results of different IoT sensing terminals at the same or adjacent network positions within the same statistical time window is calculated to obtain the observation consistency index. The time interval between the current statistical time and the timestamp of the most recent observation record of the time series is calculated to obtain the observation time gap index. According to the terminal observation reliability index, the observation consistency index can be corrected by weighting, so that the observation results of terminals with higher reliability have a greater impact on the consistency assessment. The observation frequency index, the observation consistency index, and the observation time gap index together constitute the evolution index data of the candidate evolutionary network of lawn weeds within the current statistical time window. The evolution index data is used to reflect whether the weed observation on the network is sufficient, whether the observations of different terminals are stable and consistent, and whether there is a risk interval that has not been observed for a long time.

[0032] In this embodiment S2, for each candidate evolutionary lineage of lawn weeds, a statistical time window matching the lawn maintenance cycle and weed growth cycle is set. Within each statistical time window, the time series of observation records corresponding to the lineage is statistically processed. The observation frequency of the candidate evolutionary lineage of lawn weeds is defined as the number of observation records corresponding to the lineage within the current statistical time window, which is used to characterize the number of times the lineage is actually observed by the IoT sensing terminal within the time window. The consistency index of observation results is defined as the ratio of the number of records in the current statistical time window in which different IoT sensing terminals output the same weed observation results at the same or adjacent lineage locations to the total number of corresponding comparison records. In the calculation, different weights can be assigned to the results of different terminals according to the reliability of terminal observation, which is used to characterize the stability of the observation results of multiple terminals. The observation time gap index is defined as the time interval between the statistical time and the timestamp of the most recent observation record of the candidate evolutionary lineage of lawn weeds, which is used to characterize how long the lineage has not obtained new observation data.

[0033] In this embodiment S2, within a given statistical time window, observation records whose timestamps fall within the window range are selected from the time series of network observation records, and the observation frequency under this time window is obtained by directly counting the number of records. Within the same time window, the weed observation results corresponding to the same network position or merged into the same location unit according to the distance along the line by different IoT sensing terminals are compared in groups. Record pairs with completely consistent observation results are counted as consistent records, and record pairs with inconsistent results are counted as inconsistent records. The ratio of the number of consistent records to the total number of record pairs is calculated. If necessary, the record pairs are weighted according to the terminal observation reliability to obtain the observation result consistency index. Outside the statistical time window, the timestamp of the most recent observation record of the candidate evolution network of lawn weeds is read, and the time difference between the current statistical time or task scheduling time and the timestamp is used as the observation time gap index. The observation frequency, consistency and gap index can be smoothed or averaged in multiple statistical periods to form an evolution index data sequence for task decision-making.

[0034] In this embodiment S2, the evolution index data is used in the generation of cross-sensing tasks as follows: For each candidate evolutionary lineage of lawn weeds, under the current statistical time window, when the consistency index of the observation results is lower than the consistency threshold, the lineage location range corresponding to the consistency threshold is identified as the key interval where the observation results are inconsistent; when the observation time gap index exceeds the gap threshold, the interval near the most recent observation position on the corresponding candidate evolutionary lineage of lawn weeds is identified as the gap interval that has not been observed for a long time; taking the key interval and the gap interval as the target observation interval, multiple IoT sensing terminals with coverage capabilities and differences in terminal type or installation location are selected from the set of responsible terminals, and cross-sensing tasks are generated by combining the observation time constraints in the lineage terminal association data, so that the cross-sensing task is triggered only in the lineage interval where the observation frequency is insufficient, the observation results are unstable, or there is a long gap, thereby achieving accurate triggering based on evolution index data; the evolution index data is used in the adjustment of observation tasks as follows: when a candidate evolutionary lineage of lawn weeds is in a continuous multiple When the observation frequency index is consistently below the preset lower limit within a statistical time window, the network is marked as an under-observed network. The execution cycle of the observation tasks in the corresponding responsible terminal set is shortened or additional routine observation tasks are added to improve the observation coverage of the network. When the observation frequency index of a candidate evolutionary network of lawn weeds is higher than the preset upper limit and the consistency index of the observation results is consistently higher than the consistency threshold and the observation time gap index is consistently lower than the gap threshold within multiple consecutive statistical time windows, the network is marked as a sufficiently observed and stable network. The execution cycle of the observation tasks in the corresponding responsible terminal set can be appropriately extended or some redundant observation tasks can be reduced to reduce the consumption of observation resources. When a network exhibits normal observation frequency but the consistency index of the observation results is close to the threshold multiple times within a statistical time window, the observation robustness of the network can be improved by adjusting the composition of the responsible terminal set or adding a small number of supplementary observation tasks without immediately triggering cross-sensing tasks. This allows the configuration of observation tasks to be dynamically optimized as the evolution index data changes.

[0035] In this embodiment S2, the cross-sensing task is an observation task generated based on the evolutionary index data for the target observation interval where the consistency index of the observation results in the candidate evolutionary lineage of lawn weeds is lower than a preset threshold and the observation time window index is higher than a preset threshold. The cross-sensing task includes the lawn weed candidate evolutionary lineage identifier, the target observation interval parameter, the identifier of the participating IoT sensing terminal, and the cross-observation time window parameter, which is used for multiple IoT sensing terminals to perform supplementary observations within the same target observation interval in the cross-observation time window.

[0036] In this embodiment S2, a cross-sensing task is generated for candidate evolutionary networks of lawn weeds based on evolutionary index data. When the observation consistency index is lower than a preset consistency threshold or the observation time window index is higher than a preset window threshold, the corresponding network location range is marked as the target observation interval. The target observation interval is represented by the start and end parameters of the distance along the line or the node interval parameters. For each target observation interval, the system selects multiple IoT sensing terminals from the set of responsible terminals that have an observable range in the target observation interval and have different terminal types or horizontal distances between their installation locations greater than a preset difference threshold as participating terminals. The system also calculates the common available time interval of each participating terminal in the target observation interval by combining the observation time constraints in the network terminal association data. The common time interval is extracted as a cross-observation time window. Based on this, a cross-sensing task object is generated, which includes candidate evolutionary context identifiers for lawn weeds, target observation interval parameters, a list of participating terminal identifiers, and cross-observation time window parameters. The cross-sensing task is written into the task queue and distributed to each participating terminal for execution before the start of the cross-observation time window. The participating terminals complete the observation of the target observation interval within the cross-observation time window and transmit the observation results back. The system uniformly writes the observation records generated by the cross-observation into the context observation record time series, and updates the terminal observation reliability index of the corresponding terminal based on the consistency of the observation results of multiple terminals, thereby continuously correcting the assessment of the terminal observation reliability while maintaining the integrity of the time series.

[0037] In this embodiment S2, the configuration of the consistency threshold and the gap threshold is based on the lawn maintenance cycle, weed growth cycle, and on-site management strategy. The consistency threshold represents the minimum consistency ratio required for the observation results of multiple terminals to be considered stable and consistent within the statistical time window. When the observation consistency index is lower than the consistency threshold, it indicates that there is a significant discrepancy in the observation of the same lineage location by different terminals within the current statistical time window, and supplementary verification is required through cross-sensing tasks. The gap threshold represents how long a candidate evolution lineage of lawn and weeds should not be observed by any terminal before it should be considered as having an observation gap. When the observation time gap index exceeds the gap threshold, it indicates that the corresponding lineage lacks observation data for a long period of time, and it is necessary to re-obtain the latest observation information through cross-sensing tasks. The consistency threshold and the gap threshold can be configured in a graded manner according to the lawn area type, season, and maintenance level. Higher consistency requirements and stricter gap restrictions are adopted in high-risk or high-value areas, while relatively lenient threshold settings are adopted in ordinary areas to form an adjustable balance between observation resource consumption and monitoring accuracy.

[0038] In this embodiment S2, after the cross-sensing task is completed, the terminal observation reliability index is updated through the observation results of multiple terminals. Within the target observation interval, based on the observation results of multiple IoT sensing terminals participating in the cross-sensing, a reference observation result is determined by majority voting or a weighted voting method combining existing reliability indices. Observation records consistent with the reference observation results are marked as consistent observations, and observation records that deviate from the reference observation results are marked as deviation observations. The number of consistent observations and the number of deviation observations for each terminal are accumulated and counted within multiple cross-sensing cycles, and the terminal observation reliability index is dynamically adjusted accordingly. The observation reliability index of terminals with a higher proportion of consistent observations is gradually increased, while the observation reliability index of terminals with a higher proportion of deviation observations is gradually decreased. When the observation reliability index of a certain terminal drops below the preset reliability threshold, the weight of that terminal in the comprehensive score is reduced in the candidate set of responsible terminals, or the terminal is no longer selected into the set of responsible terminals. This reflects the impact of long-term observation behavior on terminal reliability in a statistical sense.

[0039] In this embodiment S3, the evolutionary feature data is a set of feature parameters calculated based on the time series of the network observation records within multiple observation periods. It is used to characterize the first appearance location of weeds on the candidate evolutionary network of lawn weeds, the spread trend along the network direction, and the recurrence at local network locations. It is also used to distinguish between stable weed growth and spread sections and high-incidence and easily spread sections of weeds on the candidate evolutionary network of lawn weeds. The data structure of the evolutionary feature data is specifically a set of feature records that use the lawn weed candidate evolutionary network identifier as an index and record the first appearance location parameter, the spread direction parameter, and the recurrence frequency parameter within each preset statistical time window. The first appearance location parameter is used to indicate the network location interval where the weed is first observed on the candidate evolutionary network of lawn weeds. The spread direction parameter is used to indicate the displacement direction and displacement amplitude of the weed observation location along the network direction within a continuous observation period. The recurrence frequency parameter is used to indicate the number of times the weed is observed to reappear at the same network location or adjacent network locations within a preset time range.

[0040] In this embodiment S3, for each candidate evolutionary lineage of lawn weeds, evolutionary feature data is constructed based on the time series of lineage observation records. In the time dimension, the entire observation timeline is divided into statistical time windows that match the lawn maintenance cycle and weed growth cycle. In the spatial dimension, the candidate evolutionary lineage of lawn weeds is discretized into several positional units according to preset distance intervals or topological node spacing, with each positional unit corresponding to a continuous interval on the lineage. Within each statistical time window, observation records that fall into the window and whose positions fall into each positional unit are aggregated. First, records that change from a "no weeds" observation state to a "weeds" observation state are searched in time sequence. The minimum and maximum positions along the line of the set of positional units where the first change occurs are used as the first occurrence positional parameters under that window, which are used to characterize the weeds in that lawn weed. The initial occurrence interval on the candidate evolutionary lineage is determined. Subsequently, the weighted center position of the weed observation records within each statistical time window is calculated according to the location unit. The displacement direction and displacement amplitude of the weighted center position between two adjacent statistical time windows are combined and recorded as the propagation direction parameter, which is used to characterize the directionality and intensity of weed propagation along the candidate evolutionary lineage of lawn weeds. Within the statistical time range, the number of occurrences of the "weed present" observation state in each location unit and its neighboring location units, as well as the number of times it recurs in different statistical time windows, are accumulated to obtain the recurrence frequency parameter, which is used to characterize the frequency of weed recurrence in that location interval. The evolutionary feature data uses the lawn weed candidate evolutionary lineage identifier as an index, and organizes the first occurrence location parameter, propagation direction parameter, and recurrence frequency parameter corresponding to each statistical time window in chronological order as a feature record set.

[0041] In this embodiment S3, the specific rules for screening the network segment are as follows: On a given candidate evolutionary network of lawn weeds, when the first occurrence position parameter within the network position interval falls within a preset length range, and the displacement direction of the weed observation position characterized by the propagation direction parameter remains consistent along the network direction within multiple consecutive statistical time windows, and the recurrence number parameter exceeds a preset recurrence threshold, the network position interval is determined as a key evolutionary network segment; if multiple adjacent position intervals satisfy the above conditions and the distance between them is lower than a preset distance threshold, then the adjacent position intervals are merged to form a key evolutionary network segment with a larger coverage area.

[0042] In this embodiment S3, the determination of key evolutionary segments is based on evolutionary feature data and segment selection rules, performing interval screening and interval merging processing. On each candidate evolutionary line of turfgrass weeds, the position along the line is divided into several candidate position intervals. For each candidate position interval, it is checked whether the first occurrence position parameter falls within that interval and whether the interval length does not exceed a preset length threshold. It is also checked whether the sign of the propagation direction parameter remains consistent across multiple consecutive statistical time windows or only fluctuates in the opposite direction in individual windows. Finally, it is checked whether the recurrence count parameter exceeds a recurrence threshold. When all three conditions are met, the candidate position interval is marked as the initial evolutionary segment. Initial key segments are merged into larger key evolutionary segments. During the merging process, recurrence frequency and propagation direction parameters are summarized or weighted to reflect the overall propagation intensity and trend of the merged segments. The key evolutionary segments are represented in the data structure by candidate evolutionary network identifiers for lawn weeds and start and end position parameters along the line. In subsequent processing, these parameters are used to determine the spatial target objects generated by the network eradication task. At the same time, they can be used as statistical units in the effect evaluation stage to compare the changes in evolutionary feature data before and after eradication.

[0043] In this embodiment S3, the task configuration parameters are a set of parameters determined based on the evolutionary feature data corresponding to the key evolutionary network segment and the IoT operation status data associated with the spatial location of the key evolutionary network segment. These parameters include task type parameters, task execution time parameters, task priority parameters, and resource constraint parameters for the key evolutionary network segment. The network management task list is a set of task records indexed by the key evolutionary network segment identifier. Each task record includes the key evolutionary network segment identifier, the network location parameters of the key evolutionary network segment, the corresponding task configuration parameters, and the target management terminal type information.

[0044] In this embodiment S3, the IoT operational status data associated with the spatial location of the key evolutionary segment includes sprinkler irrigation plan information recorded by the irrigation control subsystem, fertilization and pruning operation plan information recorded by the maintenance management subsystem, environmental information such as wind speed, temperature, and precipitation forecasts recorded by the environmental monitoring subsystem, and operational information such as the online status of the pest control terminal equipment, available operation time window, remaining pesticide amount, and reachable area range recorded by the pest control terminal management subsystem. The key evolutionary segment and the IoT operational status data are linked through spatial location and time interval: spatially, the position parameters along the key evolutionary segment are associated with the sprinkler head position. The system spatially overlays the location, work zone boundaries, and terminal reachability to determine which sprinkler units, work zones, and treatment terminals affect the key evolutionary segment. In the time dimension, it performs time intersection calculations with the predetermined candidate time intervals for treatment execution, the sprinkler schedule, the fertilization and pruning schedule, and the available time windows for terminal operations to determine the set of time windows where treatment operations can actually be performed for each key evolutionary segment. This spatial-temporal dual-dimensional correlation between IoT operational status data and key evolutionary segments provides an operational environment basis for task type parameter selection, task execution time parameter determination, and resource constraint parameter setting.

[0045] In this embodiment S3, the task configuration parameters are determined based on the evolutionary characteristic data of key evolutionary segments and the IoT operation status data. When the recurrence frequency parameter of a key evolutionary segment is high and the propagation direction parameter indicates continuous diffusion towards important areas of the lawn or densely populated areas, the task type parameter is set to a spraying along the line or a patch spraying mode, and the task priority parameter is set to a high level. When the first occurrence location parameter of a key evolutionary segment is concentrated and the propagation direction parameter changes little, and the recurrence frequency parameter is moderate, the task type parameter is set to a spot spraying or local mechanical mowing mode, and the task priority parameter is set to a medium level. When the IoT operation status data indicates that the relevant sprinkler unit is in a closed state for certain periods of time and the environmental monitoring information indicates that the wind speed and temperature are both within a suitable range, the task execution time parameter is set to a time when the above environmental conditions are favorable and not... Within time windows that conflict with sprinkler irrigation plans and other maintenance schedules; in terms of resource constraints, the maximum length of the key evolutionary network segment that a single task can cover and the upper limit of the number of terminals that can be used in parallel are set according to the remaining amount of pesticides, available operation time, and reachable area length of the target treatment terminal equipment; the network management task list uses the key evolutionary network segment identifier as the primary key, and generates one or more task records for each key evolutionary network segment. Each task record includes the start and end position parameters along the key evolutionary network segment, task type parameters, task execution time parameters, task priority parameters, resource constraint parameters, target management terminal type information, and task status field. The task status field is used to mark the task as being in the state of pending issuance, issued but not executed, executing, or completed. The network management task list serves as the basic data structure for subsequent treatment terminal scheduling and treatment effect tracking.

[0046] In this embodiment S3, the specific method for generating the network management task list is as follows: For each key evolutionary network segment, read the IoT operation status data associated with the spatial location of the key evolutionary network segment, combine the evolutionary feature data to determine the appropriate task type parameters and task execution time parameters, assign task priority parameters to the key evolutionary network segment according to the weed propagation degree and recurrence situation represented by the evolutionary feature data, and determine the resource constraint parameters and target treatment terminal type according to the IoT operation status data. Assemble the key evolutionary network segment identifier, network location parameters and corresponding task configuration parameters to form a network treatment task record and add it to the network management task list.

[0047] In this embodiment S4, the disease control terminal equipment includes backpack spraying devices, vehicle-mounted spraying terminals, tracked or rail-mounted spraying robots, and mobile work implements with spraying functions. All of these are pre-registered in the system with their terminal type, reachable area, spraying width, minimum turning radius, and available operating time. Based on the key evolutionary network segment identifiers, network location parameters, task type parameters, task execution time parameters, resource constraint parameters, and target management terminal type information in the network management task list, the disease control terminal equipment that is available within the corresponding execution time window and can reach the spatial location of the key evolutionary network segment is selected. The start and end positions of the key evolutionary network segment are converted into a sequence of executable path points on the lawn structure topology map, and combined with the task type parameters, a path including the travel path and travel speed is generated. Terminal control commands, including those for controlling spray intensity, spray switch, spray flow rate, and spray width, are sent to the corresponding weed control terminal equipment via the terminal communication interface. The terminal equipment then moves along the trajectory of key evolutionary segments according to the control commands and performs spraying or mechanical mowing operations. During the operation, it transmits its current location, operational status, and spraying execution status. The management platform determines whether the key evolutionary segment has been completely controlled based on the path point coverage of the key evolutionary segment and updates the task status of the corresponding segment's weed control task. Key evolutionary segments reflect the areas where weeds concentrate and spread along structural lines such as irrigation lines, drainage paths, and weed trails. Continuously controlling the weed control terminal equipment along these segments helps cover recurrence areas and spread paths while reducing intervention in unaffected lawn areas.

[0048] Example 2: The present invention proposes an IoT-sensing lawn management system, which is applied to the IoT-sensing lawn management method proposed in Example 1. It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the IoT-sensing lawn management method in Example 1.

[0049] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A method for precise weed control and management of lawns using Internet of Things (IoT) sensing, characterized in that, Includes the following steps: S1. Acquire lawn structure data and IoT sensing terminal data, construct a lawn structure topology map based on the lawn structure data, extract structural line elements from the lawn structure topology map to construct a candidate evolutionary network of lawn weeds, calculate the observable relationship between the candidate evolutionary network of lawn weeds and the IoT sensing terminal based on the IoT sensing terminal configuration data, and generate network terminal association data. Among them, the candidate evolutionary network of lawn weeds is a linear topological object that describes the relationship of the growth and spread paths of weeds distributed along the lawn structure line within the lawn area, and is used to define the observation range and operation range of IoT sensing terminals. In S1, the lawn structure topology map is a structural map composed of lawn area boundaries, internal facility locations, and walking path nodes as topological nodes, and the lines connecting the nodes as topological edges. It is used to characterize the connection and spatial relationships between lawn area boundaries, sprinkler systems, drainage systems, and walking paths. The structural line elements are sprinkler pipeline routing lines, drainage runoff lines, walking path centerlines, and lawn-hard paving boundary lines extracted from the lawn structure topology map. These serve as basic linear data for the evolutionary trajectory of lawn weed candidates. The structural line elements include sprinkler pipeline routing lines, drainage runoff lines, walking path centerlines, and lawn... The candidate evolutionary network of lawn weeds is also used as an index unit for the time series of tissue network observation records, and as an association identifier between the set of responsible terminals, key evolutionary network segments and the network eradication task list; the construction method of the candidate evolutionary network of lawn weeds is as follows: based on the lawn structure topology map, the structural line elements are classified, the structural line elements adjacent to sprinkler irrigation facilities, drainage facilities and walking paths are selected, the selected structural line elements are spliced ​​according to the topological connectivity relationship to form connected lines, and a unique network identifier is assigned to each connected line. The connected lines with network identifiers are used as candidate evolutionary networks of lawn weeds. In S1, the observable relationship between the candidate evolutionary network of lawn weeds and the IoT sensing terminal is determined based on the installation location, field of view, installation height, and effective observation distance of the IoT sensing terminal. This relationship defines the observable interval and corresponding observation conditions of the candidate evolutionary network of lawn weeds within the field of view of the IoT sensing terminal, and is used to limit the observable segment of the IoT sensing terminal on each candidate evolutionary network of lawn weeds. The network terminal association data is generated based on the observable relationship and records the identifier of the candidate evolutionary network of lawn weeds, the identifier of the IoT sensing terminal, the start and end positions of the observable interval of the IoT sensing terminal on the corresponding candidate evolutionary network of lawn weeds, and the observation time constraints. This data serves as the basis for generating the cross-sensing task. S2. Based on the network terminal association data, determine the set of responsible terminals for the candidate evolutionary network of lawn weeds, generate the network observation record time series and calculate the evolution index data, generate the cross-sensing task of the candidate evolutionary network of lawn weeds according to the evolution index data, and perform the cross-sensing task through the Internet of Things sensing terminal to identify the observed lawn weeds and update the network observation record time series. Among them, the network observation record time series is a data sequence that records the IoT sensing terminal identifier, observation location and weed observation results of candidate evolutionary networks of lawn weeds at each observation time in chronological order. S3. Calculate evolutionary feature data based on the time series of the network observation records, select key evolutionary network segments that meet the network segment screening rules, read the IoT operation status data associated with the spatial location of the key evolutionary network segments, determine the removal and treatment configuration parameters of the key evolutionary network segments, and generate a network removal and treatment task list. Among them, the key evolutionary network segment is the network interval with continuous weed growth and spread along the network direction identified on the candidate evolutionary network of turfgrass weeds. S4. Assign treatment terminal equipment to each key evolutionary segment according to the task list of the network treatment, and control the treatment terminal equipment to carry out treatment operations along the key evolutionary segments.

2. The method for precise control and management of lawn weeds using IoT sensing according to claim 1, characterized in that: In S2, the responsible terminal set is a subset of terminals selected from IoT sensing terminals that have observable intervals with the candidate evolutionary lineage of lawn weeds based on the network terminal association data. It includes a main observation terminal covering the target observation interval of the candidate evolutionary lineage of lawn weeds and multiple collaborative observation terminals covering overlapping intervals. These terminals are used to perform observation tasks on the candidate evolutionary lineage of lawn weeds and provide terminal resources for cross-sensing tasks. The selection criteria for the responsible terminal set include the length of the observable interval of the IoT sensing terminal on the corresponding candidate evolutionary lineage of lawn weeds, the coverage ratio of the observable interval, and the terminal observation reliability index. The responsible terminal set is formed by comprehensively sorting the observable interval and the observation reliability index and selecting IoT sensing terminals that meet the preset coverage conditions.

3. The method for precise control and management of lawn weeds using IoT sensing according to claim 2, characterized in that: In S2, the network observation record time series is a data sequence that records the IoT sensing terminal identifiers, observation times, observation location parameters, and weed observation results observed during the execution of the responsible terminal set and cross-sensing tasks, using the lawn weed candidate evolution network as the index and in chronological order. It is used to reflect the weed distribution status of the lawn weed candidate evolution network at different observation times and to provide input for the calculation of evolution index data and evolution feature data. The data structure of the network observation record time series is specifically a list of observation records sorted by observation time with the lawn weed candidate evolution network identifier as the primary key. Each observation record includes at least the observation timestamp, IoT sensing terminal identifier, network location parameters, and weed observation result fields. The method for generating the time series of network observation records is as follows: during the execution of observation tasks and cross-sensing tasks issued based on the set of responsible terminals, the observation data returned by the IoT sensing terminals are processed by weed identification and then sequentially added to the observation record list of the corresponding lawn weed candidate evolution network, and the observation records are sorted and updated according to the observation timestamp. Evolutionary index data is a set of statistical indicators calculated based on the time series of network observation records within a preset statistical time window. The evolutionary index data includes the observation frequency of candidate evolutionary networks of lawn weeds, the consistency index of observation results, and the observation time window index, which are used to guide the generation of cross-sensing tasks and the adjustment of observation tasks. The calculation method of evolutionary index data is as follows: within the preset time window, the number of observation records in the network observation record time series is counted to obtain the observation frequency index; the consistency ratio of weed observation results of different IoT sensing terminals at the same or adjacent network positions within the same time window is counted to obtain the observation consistency index; the time interval between the current time and the timestamp of the most recent observation record is calculated to obtain the observation time window index; and the observation frequency index, the observation consistency index, and the observation time window index are combined to form the evolutionary index data.

4. The method for precise control and management of lawn weeds using IoT sensing according to claim 3, characterized in that: In S2, the cross-sensing task is an observation task generated based on the evolutionary index data for the target observation interval where the consistency index of the observation results in the candidate evolutionary lineage of lawn weeds is lower than a preset threshold and the observation time window index is higher than a preset threshold. The cross-sensing task includes lawn weed candidate evolutionary lineage identifier, target observation interval parameters, participating IoT sensing terminal identifiers and cross-observation time window parameters, which are used to perform supplementary observations by multiple IoT sensing terminals of different terminal types and installation locations within the same target observation interval within the cross-observation time window.

5. The method for precise control and management of lawn weeds using IoT sensing according to claim 4, characterized in that: In S3, the evolutionary feature data is a set of feature parameters calculated based on the time series of the network observation record within multiple observation periods. It is used to characterize the location of the first appearance of weeds on the candidate evolutionary network of lawn weeds, the spread trend along the network direction, and the recurrence at local network locations. It is also used to distinguish between stable weed growth and spread sections and high-incidence and easily spread weed sections on the candidate evolutionary network of lawn weeds. The data structure of the evolutionary feature data is specifically a set of feature records that use the candidate evolutionary lineage identifier of lawn weeds as an index and record the first occurrence location parameter, propagation direction parameter, and recurrence number parameter within each preset statistical time window.

6. The method for precise control and management of lawn weeds using IoT sensing according to claim 5, characterized in that: In S3, the specific rules for selecting the network segment are as follows: On a given candidate evolutionary network of lawn weeds, when the first occurrence position parameter within the network position interval falls within a preset length range, and the displacement direction of the weed observation position characterized by the propagation direction parameter remains consistent along the network direction within multiple consecutive statistical time windows, and the recurrence number parameter exceeds a preset recurrence threshold, the network position interval is determined as a key evolutionary network segment; if multiple adjacent position intervals satisfy the above conditions and the distance between them is lower than a preset distance threshold, then the adjacent position intervals are merged to form a key evolutionary network segment with a larger coverage area.

7. The method for precise control and management of lawn weeds using IoT sensing according to claim 6, characterized in that: In S4, the treatment configuration parameters are a set of parameters determined based on the evolutionary feature data corresponding to the key evolutionary segment and the IoT operation status data associated with the spatial location of the key evolutionary segment. These parameters include treatment method parameters, treatment execution time parameters, treatment priority parameters, and resource constraint parameters for the key evolutionary segment. The network treatment task list is a set of task records indexed by the key evolutionary segment identifier. Each task record includes the key evolutionary segment identifier, the network location parameters of the key evolutionary segment, the corresponding treatment configuration parameters, and the target treatment terminal type information.

8. An IoT-sensing lawn weed control management system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The processor executes a computer program to implement the IoT-sensing method for precise control and management of lawn weeds as described in any one of claims 1-7.

Images