Sanitation cleaning performance evaluation method and system based on operation track recognition

Abstract

The invention discloses an environmental sanitation cleaning performance evaluation method and system based on work trajectory recognition, and relates to the technical field of intelligent environmental sanitation work management and performance evaluation, and the method comprises the steps: collecting the multi-source trajectory data of environmental sanitation vehicles and personnel, and generating a standardized trajectory data set through time synchronization, abnormality elimination and space registration; on the basis of the standardized trajectory data set, distinguishing an effective operation section and a non-operation section in combination with operation types, and labeling abnormal behaviors to output a recognition trajectory; carrying out overlay analysis on the recognition track and an electronic map, generating a regional operation completion degree, and outputting a space coverage and anomaly statistical result; and calculating a comprehensive performance score according to a regional operation completion degree result. According to the method, the effective operation section and the abnormal behavior are recognized in combination with the operation type, and the operation authenticity judgment precision is improved; and the identification track and an electronic map are superposed to calculate the work completion degree and generate a spatial statistical result, and finally a quantitative performance score is formed, so that automatic evaluation and refined management of the whole process of environmental sanitation work are realized.

Description

Technical Field

[0001] This invention relates to the field of intelligent sanitation operation management and performance evaluation technology, specifically to a sanitation cleaning performance evaluation method and system based on operation trajectory recognition. Background Technology

[0002] With the deepening of smart city and sanitation informatization construction, sanitation operation management is gradually shifting from manual inspection and paper-based assessment to data-driven intelligent supervision. In recent years, the integrated application of BeiDou / GPS positioning technology, Internet of Things communication, and Geographic Information System (GIS) has enabled the continuous collection and real-time monitoring of the operating trajectories of sanitation vehicles and personnel, providing a foundation for the quantitative analysis of sanitation operation behavior. Some regions have established operation supervision systems based on trajectory playback, achieving preliminary statistics on operation routes, durations, and area coverage. However, existing systems mostly remain at the level of single-location recording and static assessment, lacking in-depth analysis of operation behavior identification, operation standardization evaluation, and performance quantification, making it difficult to form a data closed loop and a scientific evaluation model.

[0003] Existing sanitation operation assessment methods primarily rely on manual comparison of location trajectories with task plans, failing to dynamically identify and classify trajectory characteristics and operational behaviors. This results in trajectory data, while recording paths, failing to reflect the authenticity and standardization of operations. In most traditional systems, operational trajectories are only used as a basis for coverage determination, lacking standardized trajectory data models. Trajectory noise, drift, and data gaps are prevalent, severely impacting the accuracy and repeatability of operational results. Furthermore, existing assessment algorithms typically evaluate performance based on single indicators (such as operation time or mileage), failing to comprehensively quantify multi-dimensional factors such as operational coverage, timeliness, and anomaly control, making it difficult to accurately reflect differences between different areas and operation types. For abnormal behaviors (such as boundary violations, missed sweeps, or duplicate operations), most systems lack spatial logic-based automatic detection mechanisms, still relying on manual review, which is labor-intensive and inefficient. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is that existing environmental sanitation and cleaning operation assessment methods have problems such as lack of standardized processing of trajectory data, difficulty in accurately identifying effective operation behaviors, and lack of automated anomaly detection, as well as the problem of how to achieve comprehensive performance evaluation of operation coverage, timeliness and anomaly control based on operation trajectory recognition.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a method for evaluating sanitation and cleaning performance based on work trajectory recognition, comprising: collecting multi-source trajectory data of sanitation vehicles and personnel; generating a standardized trajectory dataset through time synchronization, anomaly removal, and spatial registration; based on the standardized trajectory dataset, distinguishing between effective work segments and non-work segments and labeling abnormal behaviors to output the identified trajectories, combining the work type; overlaying and analyzing the identified trajectories with an electronic map to generate regional work completion rates and output spatial coverage and anomaly statistics; and calculating a comprehensive performance score based on the regional work completion rate results. As a preferred embodiment of the sanitation and cleaning performance evaluation method based on operation trajectory recognition described in this invention, the multi-source trajectory data includes spatiotemporal location, operation status, equipment operation and attitude information collected through positioning modules, status sensors and CAN bus interfaces installed on sanitation vehicles, operating machinery and personnel terminals.

[0007] As a preferred embodiment of the sanitation and cleaning performance evaluation method based on operation trajectory recognition described in this invention, the spatial registration includes matching the multi-source trajectory data after cleaning with the operation task number, responsibility area code and route template.

[0008] As a preferred embodiment of the sanitation and cleaning performance evaluation method based on operation trajectory recognition described in this invention, the effective operation segment and the non-operation segment are determined by the speed change of the trajectory segment, the operation switch status and the CAN signal status; when the operation switch is on and the speed is within the set threshold range, it is determined to be an effective operation segment, otherwise it is marked as a non-operation driving segment.

[0009] As a preferred embodiment of the sanitation and cleaning performance evaluation method based on operation trajectory recognition described in this invention, the abnormal behavior includes boundary crossing, missed sweeping, repetitive operation, and unplanned operation.

[0010] As a preferred embodiment of the sanitation and cleaning performance evaluation method based on operation trajectory recognition described in this invention, the method of generating regional operation completion rate and outputting spatial coverage and anomaly statistics results includes comprehensively considering operation coverage rate, operation timeliness and anomaly control, and weighting each indicator by setting a weight ratio; the weight ratio is configured by the management end according to the operation scenario, seasonal tasks and responsibility area type, and is dynamically adjusted according to historical operation data.

[0011] As a preferred embodiment of the sanitation and cleaning performance evaluation method based on operation trajectory recognition described in this invention, the calculation of the comprehensive performance score includes a hierarchical weighted analysis based on the quality of operation completion, timeliness, standardization, and anomaly control.

[0012] Another objective of this invention is to provide a sanitation and cleaning performance evaluation system based on operation trajectory recognition. This system can calculate a comprehensive performance score based on the regional operation completion results, thus solving the problems of current sanitation and cleaning operation assessment methods, such as the lack of standardized processing of trajectory data, difficulty in accurately identifying effective operation behaviors, and lack of automated anomaly detection.

[0013] As a preferred embodiment of the sanitation and cleaning performance evaluation system based on operation trajectory recognition described in this invention, the system includes: a trajectory data acquisition and preprocessing module, a trajectory recognition and behavior classification module, an operation completion and coverage calculation module, and a performance evaluation and comprehensive analysis module. The trajectory data acquisition and preprocessing module collects multi-source operation trajectory data from sanitation vehicles, personnel, and equipment terminals, and generates a standardized trajectory dataset through time synchronization, anomaly removal, and spatial registration. The trajectory recognition and behavior classification module identifies operation types based on the standardized trajectory data, distinguishes between valid and non-operation segments, and marks abnormal behaviors such as boundary crossing, missed sweeping, and repeated operations. The operation completion and coverage calculation module overlays and analyzes the identified trajectories with an electronic map, calculates the operation coverage rate, timeliness, and anomaly rate for each area, comprehensively generates operation completion results, and outputs spatial coverage and anomaly statistics. The performance evaluation and comprehensive analysis module performs hierarchical weighted analysis based on the regional operation completion results, combined with operation quality, timeliness, standardization, and anomaly control indicators, to calculate a comprehensive performance score.

[0014] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement a sanitation and cleaning performance evaluation method based on job trajectory recognition.

[0015] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a sanitation and cleaning performance evaluation method based on job trajectory recognition.

[0016] The beneficial effects of this invention are as follows: The sanitation and cleaning performance evaluation method based on operation trajectory recognition provided by this invention collects multi-source trajectory data of sanitation vehicles and personnel and performs time synchronization, anomaly removal, and spatial registration to construct a standardized trajectory dataset, solving the problems of high noise, inconsistent standards, and difficulty in reuse of traditional trajectory records. Based on the standardized trajectory dataset, it distinguishes between effective operation segments and non-operation segments and marks abnormal behaviors according to the operation type, avoiding misjudging "passing by" as "operation", and significantly improving the objectivity of operation authenticity determination. Then, it overlays and analyzes the identified trajectory with electronic maps, calculates coverage, timeliness, and anomaly rate for responsibility area grids and road buffer zones, generates regional operation completion degree, and outputs spatial coverage and anomaly statistics, achieving unified quantification and spatial visualization of on-site, timely, and standardized operations. Finally, it calculates a comprehensive performance score based on the regional operation completion degree results, providing comparable and traceable quantitative assessments for individuals, vehicles, and teams, and forming an early warning and rectification closed loop in conjunction with historical results and standard thresholds. Thus, this invention simultaneously improves authenticity, accuracy, timeliness, and operability, providing an engineering-practical technical path for the refined supervision and scheduling optimization of sanitation and cleaning. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 The first embodiment of the present invention provides an overall flowchart of a sanitation and cleaning performance evaluation method based on operation trajectory recognition. Detailed Implementation

[0019] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0020] Example 1, referring to Figure 1 As an embodiment of the present invention, a method for evaluating the performance of sanitation and cleaning based on work trajectory recognition is provided, comprising: S1: Collect multi-source trajectory data of sanitation vehicles and personnel, and generate a standardized trajectory dataset through time synchronization, anomaly removal and spatial registration.

[0021] Furthermore, positioning modules, status sensors, and communication modules are installed on sanitation vehicles, operating machinery, and personnel terminals. The positioning modules utilize BeiDou / GPS dual-mode positioning units or city-level differential positioning equipment, providing real-time location information with meter-level to sub-meter accuracy in complex urban environments. Status sensors include speed, acceleration, heading angle, and operation switch sensors, capable of detecting equipment movement status and operation start / stop signals. For vehicle-type operations, operating parameters such as engine speed, water pump status, and sweeping brush angle are collected via a CAN bus interface, enabling multi-dimensional monitoring of vehicle operating status.

[0022] At the start of a task, the terminal automatically receives the task number and task area information from the scheduling system and enters data sampling mode. The sampling period is generally set to 1-5 seconds, and the collected data includes timestamps, equipment numbers, latitude and longitude, speed, heading angle, and task status signals. The terminal performs preliminary verification and anomaly removal locally, filtering out obviously abnormal jump points or drift points before writing them to the local cache. All terminals maintain communication with the central server via 4G / 5G or LoRa self-organizing network. When a signal dead zone or communication interruption occurs, the system automatically activates the offline caching mechanism to temporarily store the sampled data in local storage. After the network is restored, the terminal retransmits data in batches according to the time sequence, and ensures the integrity and order of transmission through CRC check and incremental sequence numbering, thereby avoiding missing or duplicate trajectory data.

[0023] After receiving multi-source trajectory data, the central management server performs multi-level preprocessing operations. First, it synchronizes the time of all multi-source trajectory data to ensure that the sampling time of each terminal is consistent with the system's standard clock, avoiding trajectory misalignment caused by timing errors between different devices. Then, it performs time series interpolation and outlier removal, deleting drift points and stationary noise points by judging abnormal features such as sudden speed changes, coordinate jumps, and boundary crossings. The system also applies a smoothing filtering algorithm to optimize the trajectory curve, making the trajectory shape consistent with the actual operation path. For small gaps caused by short-term communication interruptions, the system performs linear interpolation based on the preceding and following speeds and headings; for long-term missing data, only intermittent marking is applied without filling in the gaps, to maintain the accuracy of the trajectory.

[0024] After the multi-source trajectory data is cleaned, the system performs task-level association based on the task plan, work route template, and responsibility area code. All trajectory points are labeled with task numbers and timestamps and spatially registered with the electronic base map in the geographic information system. The system projects the trajectory points onto the road centerline or the polygon boundary of the responsibility area. If the trajectory deviates from the preset route but remains within the road buffer zone, it is considered a valid operation; if the trajectory point exceeds the operation boundary or enters a restricted area, it is marked as an abnormal segment. For operation switch or CAN signal data, the system compares it with the trajectory timeline. Only when the operation switch is "on" and the speed is within the operation speed threshold range is the time period identified as a "valid operation segment"; otherwise, it is marked as driving or standby.

[0025] After completing the above processing, the system converts all data into a standardized structure and organizes and stores it using a unified data model. The trajectory record format includes fields such as timestamp, object number, task number, latitude and longitude, speed, heading angle, operation status, CAN parameters, and quality markers. Spatial data is stored in the form of point columns or line segments, and a composite index is established using "object number + time interval + area code + task number" as the index key to achieve fast retrieval and related queries. Simultaneously, the system generates data quality identifiers for each trajectory record, including information on time drift, intermittent transmission, and signal jitter, for subsequent weight adjustments in performance calculations or manual review.

[0026] Regarding data security, the system employs HTTPS or MQTT over TLS communication protocols for encrypted transmission and implements two-way authentication through device certificates to prevent data tampering or leakage. All data acquisition terminals collect only necessary data fields related to the task, and the reporting end de-identifies individual information to ensure privacy and security. Furthermore, the platform automatically generates audit logs for task initiation, data correction, and anomaly handling processes to ensure the traceability and credibility of performance evaluations. Finally, a standardized trajectory dataset is constructed.

[0027] S2: Based on a standardized trajectory dataset, and combined with the job type, distinguish between valid job segments and non-job segments, and label abnormal behaviors to output the identification trajectory.

[0028] Furthermore, standardized data records are retrieved from the standardized trajectory dataset. Each record contains fields such as timestamp, object number, task number, latitude and longitude, speed, heading angle, operation status, CAN parameters, and quality markers. Multi-source trajectory data is continuously reconstructed based on the timestamps to form a complete time-series trajectory curve, maintaining the same spatial coordinate reference. Subsequently, task templates and geographic information base maps are used to perform route fitting and segmentation on the trajectory path. Each trajectory segment is uniquely identified by the task number, time interval, and geographic location, achieving a one-to-one correspondence between the trajectory and the operation task.

[0029] After trajectory reconstruction, operational behavior features are extracted based on a multi-dimensional feature fusion algorithm. First, statistical analysis is performed on speed, acceleration, and rate of change of direction angle to calculate the driving stability and path repeatability of the trajectory segment. Then, combining the operation switch signal and CAN status information, it is determined whether the trajectory segment is in a "operation execution" state. If the operation switch is "on" and the speed is within the system-set operation speed threshold range, it is determined as a "valid operation segment." If the operation switch is "off" or the speed exceeds the operation limit, it is marked as a "non-operation driving segment." When operation switch jitter or short-term fluctuations in the CAN signal occur, the system uses a minimum duration threshold for status confirmation. Only when the operation status continuously exceeds the threshold value is it recognized as a true operation status, thus effectively avoiding misidentification.

[0030] During the identification process, the job type is determined by combining spatial location and operational rules. Different job categories are distinguished by matching trajectory points with responsibility areas, road types, and task attributes, such as road sweeping, watering for dust suppression, garbage collection, or inspection. When the spatial overlap between a trajectory and a specific work area exceeds a preset threshold and the corresponding work signal remains consistent, the trajectory segment is automatically classified into the corresponding job type. For abnormal trajectory segments, such as those exceeding the responsibility area boundary, failing to reach the designated work area, or repeatedly covering the same area, they are marked according to geofencing rules, and anomaly records are generated for subsequent performance correction and anomaly deduction analysis.

[0031] A behavior pattern matching mechanism is introduced on the server side to comprehensively judge the spatiotemporal continuity and motion characteristics of the trajectory. While analyzing speed changes, directional changes, and spatial path curvature, behavior templates are established using historical sample data to match and classify the current trajectory. When the similarity between the trajectory's temporal characteristics and a certain behavior template exceeds a threshold, the trajectory is confirmed to belong to the corresponding task type. For segments with uncertain identification, they are marked as "pending review" and further corrected in subsequent manual quality checks or model self-learning stages.

[0032] After classification, the recognition results for each trajectory segment are output in a structured format, recording fields such as trajectory segment number, task type, start and end time, geographical range, average speed, coverage area, and anomaly markers. All recognition results are associated with the original task number and object number and are synchronously written to the trajectory recognition result table.

[0033] S3: Overlay and analyze the identified trajectory with the electronic map to generate the regional operation completion rate and output spatial coverage and anomaly statistics results.

[0034] Furthermore, the identified trajectories are spatially overlaid with an electronic map of sanitation operations. The electronic map is divided into responsibility areas or road grids, with each work area having a unique area code consistent with the task number. A spatial matching algorithm is used to determine whether the spatial coordinates of a trajectory segment fall within the target responsibility area, thereby counting the number of valid work trajectory points and coverage time within a unit grid. When the time a trajectory point spends within a certain area continuously exceeds a preset threshold and the corresponding work status is "valid work segment," the area is considered to have completed valid work coverage. For segments of repeated work, the system only counts coverage once during statistics; for missed or out-of-bounds areas, they are separately marked and added to the anomaly list.

[0035] After spatial matching is completed, the system calculates the task coverage rate and completion rate. Task coverage rate reflects the proportion of the task area covered by the effective trajectory, while task completion rate reflects the degree of conformity between the actual task execution and the planned task, expressed as:

[0036] in, Indicates the first A comprehensive score for the completion of tasks in each work area; Indicates the first The area within a work area covered by the effective trajectory; Indicates the first The total planned operating area for each work zone; Indicates the first The total effective working time of equipment or personnel within each work area; Indicates the first Planned work duration for each work area; Indicates the first The percentage of abnormal operations in each work area (including the percentage of abnormal behaviors such as crossing boundaries, missing sweeps, and duplicate operations); The weighting coefficients for coverage, timeliness, and anomaly rate are respectively, satisfying... It can be dynamically adjusted according to management strategies.

[0037] It should be noted that in this invention The weighting coefficients for the three indicators—coverage, timeliness, and anomaly control—are used to control the relative influence of each indicator in the overall completion calculation. To ensure the scientific rigor, stability, and traceability of the calculation results, this invention designs a weighting determination mechanism based on rule initialization and dynamic adaptation, ensuring that it not only meets management objectives but also automatically adjusts according to actual operational conditions.

[0038] During the system deployment phase, the management system sets an initial weight template based on the type of sanitation operation, the characteristics of the responsibility area, and seasonal task requirements. This template is predefined in the system configuration interface and is used to guide the initial scoring behavior of the calculation module. For example, in routine cleaning operations, to highlight the completion rate and area coverage quality, the system assigns a weight to the coverage rate. Set to 0.50, timeliness weight Set to 0.30, exception control weight Set to 0.20; however, for special tasks (such as high-temperature dust suppression and nighttime cleaning), to enhance time response and task execution speed, the value can be appropriately increased. The proportion. All weight templates satisfy the constraint that "non-negative and sum equals 1", and are written to the weight template database after configuration, forming a correspondence with the task number and responsibility area code.

[0039] Secondly, after the system has been running continuously and accumulated a certain amount of high-quality data, the platform performs experience calibration based on historical work results and on-site inspection records. The system periodically performs statistical analysis on task samples from the previous period (e.g., one month or one quarter), comparing the deviation between the work completion score calculated by the system and the results of manual assessment. If it is found that the contribution of a certain indicator (e.g., coverage rate) to overall quality is underestimated or overestimated, the platform automatically adjusts the corresponding weight ratio to make the results more consistent with on-site management understanding. This adjustment process is executed in the background, and the adjustment range is limited to preset thresholds (e.g., a single change does not exceed ±10%) to prevent short-term fluctuations from causing drastic fluctuations in performance evaluation.

[0040] A dynamic correction mechanism is introduced to continuously monitor the stability of each indicator during actual operation. When the volatility of a certain indicator exceeds a threshold in the task samples across multiple consecutive periods (e.g., the variance of the coverage rate exceeds 1.2 times the historical average), the system automatically triggers a weight adjustment suggestion, recalculating the influence ratio of the three indicators based on the latest data distribution. The adjustment result takes effect only after being reviewed and confirmed by the management side, and the system automatically generates a new version weight number and pushes it to the calculation module. The version number is bound to the task number to ensure the traceability of all calculation results.

[0041] After each weight adjustment, a weight version table is automatically generated, recording the version number, effective time, applicable task type, and modification description. The corresponding version weight is automatically applied during execution to ensure consistency and comparability of evaluation results for the same task at different time periods. Through this mechanism, this invention achieves refined management of task completion calculation weights, enabling... The value of can reflect policy guidance and adaptively reflect changes in the work site, ensuring that the overall completion result is true, robust and reproducible in the long term.

[0042] Scoring based on the work area for each work object (vehicle or personnel). Generate the regional task completion result. If a certain region If the value falls below the set threshold of 0.8, an alert is automatically triggered and the area is added to the anomaly review list. For areas with a high anomaly rate, the system further traces the trajectory identification records, locates the problematic segments, and analyzes the causes of the anomalies (such as route deviation, excessive downtime, or equipment malfunction).

[0043] Furthermore, to ensure data reliability, the calculation module of this invention maintains a unified data indexing mechanism with the previous two steps, using "object number + task number + time interval + area code" as the association key to directly retrieve original trajectory segments from the trajectory database, thereby automating and ensuring traceability of the calculation process. The system supports multi-dimensional summary statistics by shift, task, and day / month, and can output coverage reports and work completion curves, providing a data foundation for subsequent performance analysis.

[0044] S4: Calculate the overall performance score based on the regional task completion results.

[0045] Furthermore, based on the generated job completion result table, the historical task records of all job objects are grouped and summarized. Each job object is identified by a unique index of "object number + task number," and the system groups the historical task records within different task cycles. The data is aggregated over time series and supplemented with auxiliary indicators such as work duration, equipment uptime, failure rate, and energy consumption to construct a basic performance sample set. The basic performance sample set is entirely derived from standardized trajectory datasets to ensure data consistency and fairness in the evaluation.

[0046] A performance indicator system was established in accordance with sanitation service management requirements. This system includes four primary indicators: work completion quality, work timeliness, work standardization, and anomaly control level. Work completion quality is primarily determined by… The weighted average is used to reflect the coverage and execution of each responsibility area; the timeliness of the operation refers to the completion speed and delay ratio of the planned task, which is automatically calculated by the system based on the ratio of the actual operation time to the planned time; the standardization of the operation is judged based on parameters such as trajectory smoothness, route deviation rate and speed stability to determine whether the operation meets the operating standards; the level of anomaly control is evaluated by the anomaly rate and anomaly response time to assess the frequency of operation anomalies and the efficiency of handling them.

[0047] After standardizing the above four indicators, a comprehensive performance score is generated based on the weighting coefficients set by the management department, as follows:

[0048] in, Indicates the first The overall performance score of each work entity (which may be workers or vehicles); Indicates the first The task completion quality indicators for each task entity are derived from the completion rate of each region. The weighted average; Indicates the first The timeliness indicator for each work entity, namely the on-time rate of actual task completion; Indicates the first The operational standardization indicators for each work entity are used to reflect the stability of the work trajectory, the degree of route adherence, and the consistency of operation. This represents the job defect rate, and the calculation method is the same as... Consistency is used to reflect the proportion of abnormal behavior; The weighting coefficients for each indicator satisfy the following conditions: The weight values ​​can be adaptively adjusted based on different city management strategies or seasonal task characteristics.

[0049] It should be noted that after the performance results are generated, the system will calculate the overall score for each task. This is mapped to a visual management interface. The management side can view performance distribution by responsibility area, work group, or time period, and present dynamic performance trends in the form of heatmaps, radar charts, or time-series curves. If a certain work entity's... If the value continuously falls below the system's set threshold of 0.7, the platform automatically triggers an early warning mechanism, marks the object as a "performance anomaly," and associates it with data in the abnormal trajectory segment and the area of ​​responsibility, thus achieving closed-loop management from problem discovery to accountability.

[0050] In addition, the system supports historical performance comparison and trend analysis functions. This is achieved by comparing the current period's performance... By comparing performance with the average over multiple periods, the system can identify performance fluctuation ranges and trend directions, providing management with a basis for optimization decisions. For example, when the operational quality indicators of a certain responsibility area... The rate continued to decline, while the anomaly rate... When the trend is upward, the system will suggest replanning routes, adjusting equipment maintenance cycles, or optimizing personnel allocation.

[0051] In this invention These correspond to the weighting coefficients of four core indicators: task completion quality, timeliness, standardization, and anomaly control. This set of weights is used to balance the distribution among multi-dimensional performance parameters and is crucial for generating a comprehensive performance score. The key parameters. To ensure the rationality and sustainable adaptability of the performance evaluation results, this invention also designs a multi-level weight determination and adaptive update mechanism.

[0052] During the initial deployment phase, the management system establishes a default performance weight template based on the sanitation management model and assessment requirements. This template is tiered according to the type of work entity (vehicles, manual labor, machinery), the level of the work area (main roads, secondary roads, side streets, back streets), and seasonal characteristics (normal season, rainy / snowy season, special cleaning season). For example, in a normal work scenario, to emphasize work quality and standardization, the system defaults to a certain configuration. =0.40、 =0.25、 =0.20、 =0.15; however, in emergency operations or nighttime rescue missions, to emphasize efficiency and response speed, it can be... Increase to 0.35, while appropriately reducing The proportion. All template parameters are subject to constraints. And it is distributed to the computing engine through the configuration management module to ensure that the initial performance evaluation criteria are consistent.

[0053] Secondly, the system periodically performs data-driven weight recalibration after continuous operation. The platform aggregates and analyzes historical performance data quarterly, calculating the correlation between the scores of each indicator and the actual assessment results of the management department. If it is found that a certain indicator has insufficient impact on the final performance (e.g., the contribution of work standardization to the overall score is too low), the system automatically proposes adjustment suggestions based on the correlation coefficient, with the adjustment ratio not exceeding a preset threshold. This process is semi-automatic; management can view the system suggestions on the interface and manually confirm them. The confirmed weight configuration takes effect immediately. New weights are applied with a version identifier and effective date, and old weight records are retained to ensure data traceability.

[0054] During long-term operation, when the platform detects significant changes in task attributes, environmental conditions, or work scale (such as holiday support, abnormal weather, road construction detours, etc.), it automatically matches the corresponding weight template to adapt to the new work environment. For example, when the system identifies an anomaly rate in a certain responsibility area across multiple consecutive task cycles... When there is a significant increase, it will be temporarily raised. The proportion of [something] is increased to strengthen the constraint of anomaly control in performance scoring, thereby guiding operators to improve standardization.

[0055] Finally, all weight updates and version changes are recorded internally on the platform as a "Weight Configuration Version Table," which includes information such as weight refactoring number, configurer, release time, and applicable scope. The calculation module automatically reads the corresponding weight version when executing the performance scoring formula, achieving version-based management. When the management side reviews historical performance data or traces performance anomalies, it can regenerate the current performance results based on the weight version information, ensuring the consistency, accuracy, and auditability of the assessment data.

[0056] Through the aforementioned weight determination and dynamic adjustment mechanism, the present invention... This mechanism enables adaptive updates across different management stages, work types, and environmental conditions, making the overall performance results more flexible and accurate. While ensuring the sum of all weighted coefficients remains constant at 1, it also achieves stability, interpretability, and data reliability in performance evaluation, further improving the closed-loop structure of the sanitation and cleaning performance evaluation system.

[0057] Example 2, one embodiment of the present invention, provides a method for evaluating the performance of sanitation and cleaning based on operation trajectory recognition. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiment.

[0058] First, the test area selected two main roads and three secondary roads in a certain urban area, with a total responsibility area ranging from 1.4 to 4.5 ha, including typical road conditions such as straight sections, intersections, and road curves. The operation involved four sweepers and two groups of foot cleaners. The "traditional group" did not use the trajectory recognition and anomaly labeling functions of this invention, only performing routine positioning and manual check-in. The "invention group" used positioning modules, status sensors, and communication units on both vehicle and personnel terminals to collect spatiotemporal location, speed, operation status, and equipment operating information from the CAN bus. Time synchronization, anomaly removal, and spatial registration were performed on the central platform to generate a standardized trajectory dataset. The task plan and responsibility area code were issued by the scheduling system before the operation, and the terminal automatically entered the task number to ensure consistency between the trajectory record and the task description.

[0059] After the operation begins, all terminals continuously upload data at a sampling period of 1–5 seconds. When cellular signal attenuation occurs in some high-rise building sections, the terminals trigger offline caching and automatically re-upload data after communication is restored at the end of the road segment. The central platform performs time alignment, outlier removal, and trajectory smoothing on the incoming data, and spatially registers the valid records with the road centerline and responsibility area polygon in the electronic map. For the "Invention Group," the platform automatically distinguishes between valid operation segments and non-operation driving segments based on the standardized trajectory dataset, combined with the operation type and operation on / off status, and marks abnormal behaviors such as boundary crossing, missed scanning, and repeated operations, outputting the identified trajectory in real time. For the "Traditional Group," only path records and duration statistics are retained, without operation status judgment or anomaly structuring.

[0060] After the operation is completed, the platform will overlay and analyze the identified trajectory with the electronic map, statistically analyze the coverage area, effective operation time, and anomalies according to the responsibility area grid and road buffer zone, generate the regional operation completion rate, and output the spatial coverage and anomaly statistics results. Subsequently, based on the regional operation completion rate results, the platform calculates the comprehensive performance score and summarizes it hierarchically according to vehicle, personnel group, and mixed operation objects, generating a visual report and detailed list. To ensure fairness, the test was conducted under similar traffic load conditions and at the same time period; for road sections with short-term accidents and temporary road closures, the platform marked the impact range in the report and recorded it separately, and it was not included in the indicator comparison. The final data is shown in Table 1.

[0061] Table 1 Experimental Data

[0062] Observing the "effective trajectory coverage rate (%)" indicator, the traditional vehicle V1 and V2 have coverage rates of 82.4% and 79.8% respectively, while the invention-vehicle V3 and V4 reach 95.6% and 94.2% respectively. In the personnel operation scenario, the traditional personnel group P1 has a coverage rate of 73.1%, while the invention-personnel group P2 has increased to 91.8%; the mixed operation G1 reaches 96.3%. This comparison shows that the "effective operation segment" determination mechanism based on standardized trajectory datasets and operation type recognition can significantly improve the spatial performance of "operation in place", avoid counting "passing by" in operation coverage, and make the coverage statistics more consistent with the actual operation.

[0063] In comparing the effective working time with the planned time, the present invention group, while ensuring coverage, did not significantly increase the time: for example, vehicle V3 had an actual effective working time of 109 minutes under a 110-minute plan, and V4 had an actual working time of 115 minutes under a 120-minute plan. Both were closer to the planned time than the traditional group, and no abnormal overtime occurred. This indicates that job identification reduced ineffective travel segments, resulting in higher time utilization. The comparison of personnel groups also showed the same trend: P1's effective time of 87 minutes was lower than the 90-minute plan and the coverage was insufficient, while P2 achieved higher coverage within the planned 90 minutes and had an effective time of 82 minutes. This indicates that through identification and anomaly labeling, the path organization was more reasonable and redundant travel was reduced.

[0064] The anomaly rate, the number of missed grids, and the mileage of repeated coverage all reflect the standardization and anomaly control. Traditional vehicles (V1 / V2) had an anomaly rate of 6.8% / 7.5%, 23 / 31 missed grids, and 4.1 / 3.7km of repeated coverage. The present invention vehicles (V3 / V4) reduced the anomaly rate to 2.4% / 3.1%, 6 / 8 missed grids, and 1.3 / 1.6km of repeated coverage. The personnel group also showed improvement: P1 had an anomaly rate of 9.7%, 27 missed grids, and 2.2km of repeated coverage, while P2 had an anomaly rate of 3.8%, 9 missed grids, and 1.1km of repeated coverage. Hybrid operation G1, even under multi-entity collaborative conditions, maintained a low anomaly rate of 2.9%, 5 missed grids, and 1.4km of repeated coverage, indicating that the anomaly labeling and spatial statistics mechanism after overlaying the identification trajectory and electronic map can effectively suppress problems such as boundary crossings, missed grids, and disorderly repeated cleaning.

[0065] The comprehensive performance score reflects the combined results of terminal management: Traditional - Vehicle V1 / V2 is 0.71 / 0.67; Invention - Vehicle V3 / V4 is 0.92 / 0.89; Traditional - Personnel P1 is 0.61, while Invention - Personnel P2 is 0.86; Mixed Operations G1 reaches 0.93. This distribution is consistent with improvements in coverage, timeliness, and anomaly control, indicating that the closed loop of trajectory identification—spatial overlay—completion—performance score has a decisive impact on the final performance. Especially in complex road networks and high-traffic areas, this invention, by distinguishing between effective and non-operational sections and marking abnormal behaviors, enables managers to quickly locate problem road sections, identify responsible parties, and optimize route and time slot configurations based on spatial coverage and anomaly statistics.

[0066] As can be seen from the data from the examples, the present invention is significantly superior to traditional methods in terms of operational authenticity, spatial coverage accuracy, anomaly control, and performance quantification, and has clear engineering value and significance for promotion.

[0067] Example 3, an embodiment of the present invention, provides a sanitation and cleaning performance evaluation system based on operation trajectory recognition, including a trajectory data acquisition and preprocessing module, a trajectory recognition and behavior classification module, an operation completion and coverage calculation module, and a performance evaluation and comprehensive analysis module.

[0068] The system comprises several modules: a trajectory data acquisition and preprocessing module for collecting multi-source operation trajectory data from sanitation vehicles, personnel, and equipment terminals, and generating a standardized trajectory dataset through time synchronization, anomaly removal, and spatial registration; a trajectory recognition and behavior classification module for identifying operation types based on standardized trajectory data, distinguishing between valid and non-operation segments, and marking abnormal behaviors such as boundary crossing, missed sweeping, and repeated operations; an operation completion and coverage calculation module for overlaying and analyzing the identified trajectories with electronic maps, calculating the operation coverage, timeliness, and anomaly rate for each area, comprehensively generating operation completion results, and outputting spatial coverage and anomaly statistics; and a performance evaluation and comprehensive analysis module for performing hierarchical weighted analysis based on the regional operation completion results, combined with operation quality, timeliness, standardization, and anomaly control indicators, to calculate a comprehensive performance score.

[0069] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0070] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0071] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0072] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc. It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

[0073] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for evaluating the performance of sanitation and cleaning services based on work trajectory recognition, characterized in that, include: Multi-source trajectory data of sanitation vehicles and personnel are collected, and standardized trajectory datasets are generated through time synchronization, anomaly removal and spatial registration. Based on a standardized trajectory dataset, and combined with the job type, we distinguish between valid job segments and non-job segments, and label abnormal behaviors to output the identification trajectory. By overlaying and analyzing the identified trajectories with electronic maps, the regional operation completion rate is generated and spatial coverage and anomaly statistics are output. Calculate the overall performance score based on the regional task completion results.

2. The sanitation and cleaning performance evaluation method based on work trajectory recognition as described in claim 1, characterized in that: The multi-source trajectory data includes spatiotemporal location, operational status, equipment operation, and attitude information collected through positioning modules, status sensors, and CAN bus interfaces installed on sanitation vehicles, operating machinery, and personnel terminals.

3. The sanitation and cleaning performance evaluation method based on work trajectory recognition as described in claim 2, characterized in that: The spatial registration includes matching the cleaned multi-source trajectory data with the task number, responsibility area code, and route template.

4. The sanitation and cleaning performance evaluation method based on operation trajectory recognition as described in claim 3, characterized in that: The effective working segment and non-working segment are determined by a combination of speed changes in the trajectory segment, working switch status, and CAN signal status. When the work switch is on and the speed is within the set threshold range, it is determined to be a valid work segment; otherwise, it is marked as a non-work segment.

5. The sanitation and cleaning performance evaluation method based on work trajectory recognition as described in claim 4, characterized in that: The abnormal behaviors include out-of-bounds operations, missed scans, duplicate operations, and unplanned operations.

6. The sanitation and cleaning performance evaluation method based on operation trajectory recognition as described in claim 5, characterized in that: The generation of regional operation completion rate and output of spatial coverage and anomaly statistics results include comprehensive consideration of operation coverage rate, operation timeliness and anomaly control, and weighted calculation of each indicator by setting weight ratios; The weight ratio is configured by the management system based on the work scenario, seasonal tasks, and responsibility area type, and is dynamically adjusted according to historical work data.

7. The sanitation and cleaning performance evaluation method based on operation trajectory recognition as described in claim 6, characterized in that: The calculation of the comprehensive performance score includes a hierarchical weighted analysis based on the quality of task completion, timeliness, standardization, and anomaly control.

8. A system employing the sanitation and cleaning performance evaluation method based on operation trajectory recognition as described in any one of claims 1 to 7, characterized in that: It includes a trajectory data acquisition and preprocessing module, a trajectory recognition and behavior classification module, a task completion and coverage calculation module, and a performance evaluation and comprehensive analysis module; The trajectory data acquisition and preprocessing module is used to collect multi-source operation trajectory data from sanitation vehicles, personnel and equipment terminals, and generate standardized trajectory datasets through time synchronization, anomaly removal and spatial registration. The trajectory recognition and behavior classification module is used to identify the job type based on standardized trajectory data, distinguish between valid job segments and non-job segments, and mark abnormal behaviors such as exceeding boundaries, missing scans, and repeated jobs. The task completion and coverage calculation module is used to overlay and analyze the identified trajectory with the electronic map, calculate the task coverage, timeliness and anomaly rate of each area, generate the task completion result and output spatial coverage and anomaly statistics. The performance evaluation and comprehensive analysis module is used to perform hierarchical weighted analysis based on the regional operation completion results, combined with operation quality, timeliness, standardization and anomaly control indicators, to calculate the comprehensive performance score.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the sanitation and cleaning performance evaluation method based on job trajectory recognition as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the sanitation and cleaning performance evaluation method based on job trajectory recognition as described in any one of claims 1 to 7.

Images