A sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism

Abstract

The present invention relates to the technical field of sanitation service quality management, and discloses a sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism. The method processes the sanitation operation process data and service item feature data of the target area through a task decoupling model to generate a task decomposition heat map and a weight gradient field; secondly, a multidimensional evaluation index system is constructed, and a dynamic data collection strategy is formulated in combination with the DBSCAN clustering algorithm; the feedback data collection device obtains service feedback data according to the strategy, and identifies service quality levels and defective items through a random forest model; finally, based on the closed-loop feedback mechanism, the evaluation data is uploaded to the management control center for adaptive correction of indicator thresholds and weights. The present invention realizes refined modeling of sanitation operations through task decoupling, and utilizes closed-loop feedback to dynamically optimize evaluation criteria, thereby solving the problems of single evaluation dimension and lagging indicator update in traditional methods, and improving the accuracy and timeliness of sanitation service evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of sanitation service quality management, and in particular to a sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism. Background Art

[0002] Most evaluation systems treat sanitation operations as a holistic task, without deeply analyzing the characteristics of service items and operational processes, resulting in an inability to accurately identify service shortcomings. For example, traditional methods lack quantitative analysis of the correlation between service items, making it difficult to define the boundaries of different operational units. This results in overlaps or omissions in task decomposition results, which in turn affects the pertinence of evaluation indicators. At the same time, in the data collection process, traditional strategies often use fixed points or manual inspections, failing to combine task decomposition characteristics with spatiotemporal dynamics. This results in feedback data failing to fully cover high-priority evaluation areas, resulting in blind spots or redundant collection, and reducing data utilization efficiency.

[0003] Traditional evaluation index systems are typically static and lack dynamic correction mechanisms based on real-time feedback data. When sanitation operations change (e.g., seasonal fluctuations in foot traffic, sudden pollution incidents), fixed indicators fail to promptly reflect the true demand for service quality, resulting in a disconnect between evaluation results and actual management needs. For example, during peak holiday traffic periods, the weighting of cleanliness indicators continues to follow the daily standard, failing to reflect the service pressure in high-frequency operation areas, thus affecting the rationality of resource allocation.

[0004] When it comes to processing feedback data, existing technologies often use a single model to identify service quality, failing to fully consider the differences in the characteristics of different service items. This results in insufficient accuracy in identifying defective items. Furthermore, after feedback data is transmitted to management, there is a lack of a systematic indicator correction process, preventing the formation of a closed-loop management system of "evaluation-feedback-optimization." This severely limits the adaptability and timeliness of the evaluation system. Summary of the Invention

[0005] The purpose of the present invention is to provide a sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism to solve the problems raised in the above background technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism, the method comprising:

[0007] Acquire sanitation operation process data and service item characteristic data of the target area based on the task decoupling model, and determine the task decomposition of the target area according to the service item characteristic data;

[0008] Constructing an environmental sanitation service evaluation index system based on the operation process data, and determining a feedback data collection strategy based on the evaluation index system and task decomposition;

[0009] The feedback data collection device obtains service feedback data of the target area according to the collection strategy, identifies the service quality level and defect items of the target area according to the feedback data, and obtains service evaluation data;

[0010] The service evaluation data is transmitted to the management control center according to the closed-loop feedback mechanism for index correction.

[0011] Preferably, the task decoupling model is used to obtain the sanitation operation process data and service item characteristic data of the target area, and the task decomposition of the target area is determined according to the service item characteristic data, specifically:

[0012] Acquire the operation process data of the target area and the characteristic data of the service items in the target area based on the task decoupling model;

[0013] Adaptively classifying the service item feature data based on a workflow decomposition algorithm to obtain service item feature data with clear boundaries, introducing a service item relevance analysis algorithm, and setting initial thresholds for parameters of the service item relevance analysis algorithm;

[0014] Decomposing the feature data of the service items with clear boundaries into tasks according to the service item correlation analysis algorithm to obtain N independent service units;

[0015] Marking the process nodes of the independent service units, constructing an execution path graph for each independent service unit, and calculating the operation frequency, coverage, and resource input of each independent service unit based on the execution path graph to obtain a task decomposition feature set;

[0016] Acquire spatial location information of the task decomposition feature set according to the service item feature data, perform regional mapping of the task decomposition feature set at each location according to the spatial location information, and construct a three-dimensional task decomposition feature map;

[0017] The task weight of each position is determined according to the three-dimensional task decomposition feature map, the task decomposition heat map and task weight gradient field of the target area are constructed according to the task weight of each position, and the task decomposition status of the target area is determined according to the task decomposition heat map and task weight gradient field.

[0018] Preferably, the sanitation service evaluation index system is constructed based on the operation process data, and the feedback data collection strategy is determined based on the evaluation index system and task decomposition, specifically:

[0019] Perform multi-dimensional weight analysis on the operation process data, construct an indicator feature tensor through standardization processing, identify service quality factors of the execution signal of the operation process data based on the fuzzy comprehensive evaluation method, and generate an evaluation index system for target areas including operation frequency, cleanliness, and response time;

[0020] Performing a dimension mapping operation on the task decomposition and the evaluation index system to construct a multidimensional service quality evaluation model, and dividing the multidimensional service quality evaluation model according to a preset time window to construct M time intervals;

[0021] Obtaining service quality score information for each time interval according to the task decomposition, and clustering time intervals with similar score trends in the service quality multidimensional evaluation model based on the DBSCAN clustering algorithm to obtain clustering results;

[0022] Determining an evaluation priority of the service quality of each time interval according to the clustering result, obtaining corresponding spatiotemporal information of each time interval in the target area, and performing a service evaluation importance evaluation on each location in the target area according to the evaluation priority and the corresponding spatiotemporal information to obtain an evaluation importance score for each location in the target area;

[0023] Determining evaluation requirement information for each location in the target area according to the evaluation importance score of each location, wherein the evaluation requirement information includes whether to perform collection and collection density requirement information;

[0024] Obtaining information acquisition completeness data of the feedback data acquisition device for different areas of service, determining data collection distance information of the feedback data acquisition device for each location in the target area based on the information acquisition completeness data and the evaluation requirement information, and determining a collection point set of the feedback data acquisition device based on the data collection distance information;

[0025] A collection strategy of the feedback data collection device is determined according to the collection point set.

[0026] Preferably, the collection strategy of the feedback data collection device is determined according to the collection point set, specifically:

[0027] Obtaining the spatiotemporal coordinate information of each collection point and the initial position information of the feedback data collection device, and determining the operation blind area in the target area according to the evaluation index system;

[0028] Performing path planning on the initial position information and the spatiotemporal coordinate information of each collection point based on an ant colony algorithm, taking the operation blind area as a path planning restriction area, and outputting the shortest initial collection path for the feedback data collection device;

[0029] Acquire in real time device state change data of the feedback data acquisition device during the acquisition process according to the shortest initial acquisition path, and determine the crowd density and operation interference intensity at the real-time acquisition location of the feedback data acquisition device based on the state change data;

[0030] Acquiring anti-interference capability data of the feedback data acquisition device, wherein the anti-interference capability data includes information filtering capability data of the feedback data acquisition device for interferences of different intensities;

[0031] Obtaining recognition response time data of the feedback data acquisition device to the crowd density and the work interference intensity, and determining a collection path adjustment hysteresis amount of the feedback data acquisition device according to the recognition response time data;

[0032] Analyzing the crowd density and the intensity of the work interference at the real-time collection location of the feedback data collection device according to the anti-interference capability data and the collection path adjustment hysteresis, and determining the cumulative amount of collection path deviation of the feedback data collection device within the recognition response time;

[0033] If the accumulated amount of the acquisition path offset is less than a preset value, determining a path offset direction and an offset distance for the feedback data acquisition device to acquire data according to the shortest initial acquisition path based on the accumulated amount of the acquisition path offset, and determining an adjustment direction and an adjustment distance of the feedback data acquisition device based on the offset direction and the offset distance to obtain adjustment data;

[0034] Adjusting the shortest initial acquisition path of the feedback data acquisition device during the real-time acquisition process according to the adjustment data to obtain a first acquisition strategy;

[0035] If the accumulated amount of the acquisition path deviation is greater than a preset value, obtaining the pedestrian flow density and operation interference intensity data of the real-time acquisition path of the feedback data acquisition device, constructing an interference change map based on the pedestrian flow density and operation interference intensity data of the real-time acquisition path, and determining the interference change trend of the service evaluation area in the target area based on the interference change map;

[0036] An interpolation operation is performed on the interference change trend based on spline interpolation to determine the interference information within the preset range of the shortest initial acquisition path. The acquisition stability of the feedback data acquisition device within the preset range of the shortest initial acquisition path is determined based on the interference information and the anti-interference capability data. According to the acquisition stability, the shortest initial acquisition path section with the accumulated acquisition path offset greater than the preset value is optimized to obtain an updated acquisition path. The feedback data acquisition device performs an acquisition operation based on the updated acquisition path to obtain a second acquisition strategy.

[0037] Preferably, the feedback data collection device obtains service feedback data of the target area according to the collection strategy, identifies the service quality level and defect items of the target area according to the feedback data, and obtains service evaluation data, specifically:

[0038] Obtaining standard feedback data for different service items in the target area, and labeling the standard feedback data with service levels to obtain labeled feedback data;

[0039] Building a service quality recognition model based on a random forest algorithm, and importing the labeled feedback data into the service quality recognition model for training;

[0040] The feedback data collection device obtains service feedback data of the target area according to the collection strategy, imports the service feedback data into the trained service quality identification model to identify the service quality level, and counts the defective items of each service item to obtain service evaluation data.

[0041] Preferably, the service evaluation data is transmitted to the management control center according to a closed-loop feedback mechanism for index correction, specifically:

[0042] Building a layered data correction protocol stack based on a closed-loop feedback mechanism, extracting features from the service evaluation data, and building a service evaluation feedback signal;

[0043] The service evaluation feedback signal is transmitted to a management control center according to the layered data correction protocol stack, and an indicator mapping operation is performed on the service evaluation feedback signal to obtain the sanitation service evaluation correction data of the target area.

[0044] Preferably, the service evaluation feedback signal is transmitted to the management control center according to the layered data correction protocol stack, and the service evaluation feedback signal is subjected to an indicator mapping operation to obtain the sanitation service evaluation correction data of the target area, specifically:

[0045] Obtaining a historical evaluation database of a management control center, performing similarity matching between the service evaluation feedback signal and the historical evaluation database to obtain a matching result;

[0046] Determine the distribution characteristics of abnormal data in the service evaluation feedback signal based on the matching results, and perform a time series filtering operation on the abnormal data based on a sliding window algorithm to obtain valid feedback data;

[0047] Aligning the effective feedback data with the evaluation index system, constructing an index correction correlation matrix, and determining the evaluation index items that need to be adjusted according to the weight change trend of each index in the correlation matrix;

[0048] The threshold range or weight coefficient of the evaluation index items that need to be adjusted is corrected to generate an updated sanitation service evaluation index system, and the updated evaluation index system is stored in the management and control center to obtain the sanitation service evaluation correction data of the target area.

[0049] Preferably, the effective feedback data is dimensionally aligned with the evaluation index system to construct an index correction correlation matrix, and the evaluation index items that need to be adjusted are determined according to the weight change trend of each index in the correlation matrix, specifically:

[0050] Performing principal component analysis on the effective feedback data to extract key factors affecting service quality, and determining corresponding correlation relationships with the evaluation index system based on the key factors;

[0051] Calculate the correlation coefficients between each evaluation index and key factors based on the covariance matrix and construct the index correction correlation matrix;

[0052] The correlation coefficients of the indicators in the correlation matrix are sorted, and evaluation indicator items whose absolute values ​​of the correlation coefficients are less than a preset threshold are screened out as evaluation indicator items that need to be adjusted.

[0053] Preferably, the threshold range or weight coefficient of the evaluation index item that needs to be adjusted is modified to generate an updated sanitation service evaluation index system, specifically:

[0054] For quantitative indicators in the evaluation indicators that need to be adjusted, the upper and lower thresholds will be reset based on the statistical distribution of valid feedback data;

[0055] For qualitative indicators in the evaluation indicators that need to be adjusted, redefine the level description based on the semantic analysis results of the effective feedback data;

[0056] Normalize and adjust the weight coefficients of all evaluation index items to ensure that the total weight is 1, and generate an updated sanitation service evaluation index system.

[0057] Preferably, the construction of a multi-dimensional service quality evaluation model is specifically as follows:

[0058] Determine the core dimensions of service quality assessment, including operational standardization, cleanliness compliance rate, response timeliness, and complaint handling rate;

[0059] Specific evaluation indicators are set for each core dimension, including: operational standardization includes equipment compliance and operational process integrity; cleaning compliance rate includes road sweeping cleanliness and garbage collection and transportation timeliness; response timeliness includes emergency response time and task dispatch completion timeliness; complaint handling rate includes complaint acceptance rate and problem resolution satisfaction;

[0060] Based on the Delphi method, an expert algorithm is used to evaluate the importance of each dimension and indicator. The weight coefficient of each dimension and the sub-weight coefficient of each indicator within the corresponding dimension are calculated according to the scoring results.

[0061] The evaluation indicators, weight coefficients and sub-weight coefficients of each dimension are structurally integrated to construct a multidimensional evaluation model of service quality that includes multidimensional indicators and hierarchical weights.

[0062] Compared with the prior art, the present invention has the following beneficial effects:

[0063] The present invention significantly improves the accuracy and dynamic adaptability of sanitation service evaluation through the organic combination of the task decoupling model and the closed-loop feedback mechanism. Based on the in-depth analysis of the operation process and service item characteristics by the task decoupling model, it can realize the adaptive classification and independent unit disassembly of service items through workflow decomposition algorithm and correlation analysis, construct a three-dimensional task decomposition feature map and task weight gradient field, make the task decomposition more in line with the actual operation scenario, and lay the foundation for accurate evaluation. This refined decomposition can clearly define the operation frequency, coverage and resource input of each service unit, avoid the problem of blurred task boundaries in traditional methods, and thus accurately locate inefficient links during evaluation.

[0064] In terms of constructing an evaluation index system, a multi-dimensional weight analysis and fuzzy comprehensive evaluation method were used to generate an index system encompassing core elements such as operation frequency and cleanliness. This was then dimensionally mapped to the task decomposition. The DBSCAN clustering algorithm was then used to determine spatiotemporal evaluation priorities, making the collection strategy more targeted. This strategy, based on evaluation importance scores and information acquisition completeness, uses an ant colony algorithm to optimize and adjust the collection path in real time. This not only covers blind spots in operations but also dynamically adjusts the path based on interfering factors such as traffic density, ensuring the integrity and timeliness of feedback data. This approach better reflects the true quality of service compared to traditional fixed-point collection.

[0065] The service quality identification link uses the random forest algorithm to train the recognition model. By labeling the level of standard feedback data and training the model, it can accurately identify the service quality level and count the defect items. Compared with a single model, it has higher classification accuracy and robustness, and can effectively improve the accuracy of defect item identification.

[0066] The closed-loop feedback mechanism uses a layered data correction protocol stack to extract and transmit service evaluation data. It combines the historical evaluation database to filter abnormal data and map indicators. It uses principal component analysis and covariance matrix to construct an indicator correction association matrix, adjusts the thresholds and redefines the levels of quantitative and qualitative indicators respectively, and normalizes the weight coefficients. This process enables the evaluation indicator system to be dynamically optimized based on real-time feedback data, forming a closed-loop management of "data collection-quality assessment-indicator correction", solving the problem of insufficient adaptability of traditional static indicator systems, enabling the evaluation system to respond to changes in operating scenarios in real time, and continuously improving the scientific nature of evaluation and the effectiveness of management. In addition, this method realizes intelligent management of the entire process from task decomposition to indicator correction through multi-algorithm fusion and dynamic adjustment mechanism, providing effective support for the refined and scientific management of sanitation services, and can significantly improve resource allocation efficiency and service quality control. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 This is a working principle diagram of the sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism according to the present invention;

[0068] Figure 2 This is a design diagram of the closed-loop feedback mechanism;

[0069] Figure 3 Correction of the design diagram associated with the indicator;

[0070] Figure 4 Design diagram constructed for the incidence matrix. DETAILED DESCRIPTION

[0071] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0072] See also Figure 1-Figure 4 The present invention relates to a sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism, and the specific implementation steps are as follows:

[0073] Based on the task decoupling model, the sanitation operation process data and service item characteristic data of the target area are obtained, and the task decomposition of the target area is determined according to the service item characteristic data.

[0074] Construct an environmental sanitation service evaluation index system based on operation process data, and determine the feedback data collection strategy based on the evaluation index system and task decomposition.

[0075] The feedback data collection device obtains the service feedback data of the target area according to the collection strategy, identifies the service quality level and defect items of the target area according to the feedback data, and obtains service evaluation data.

[0076] The service evaluation data is transmitted to the management control center according to the closed-loop feedback mechanism for indicator correction.

[0077] Example 1: After obtaining the sanitation operation process data and service item feature data of the target area based on the task decoupling model, it is necessary to determine the task decomposition of the target area according to the service item feature data. The specific implementation method is as follows:

[0078] Based on the task decoupling model, we acquire operational process data for the target area. This data covers all aspects and steps of sanitation operations within the target area, such as the specific operation methods and time schedules for cleaning, garbage collection, and transportation. We also acquire feature data for service items within the target area, including information on their type, nature, complexity, and required resources.

[0079] Adaptively classify service item feature data based on the workflow decomposition algorithm. This algorithm automatically categorizes service item feature data based on its characteristics, ensuring clear boundaries and similar characteristics within each category. During classification, the algorithm analyzes various attributes of the service item feature data, such as type and complexity, and groups services with similar attributes into a single category, resulting in clearly defined service item feature data.

[0080] Introduce a service item correlation analysis algorithm and set initial thresholds for its parameters. The service item correlation analysis algorithm analyzes the correlation between services. By setting initial thresholds, you can filter out services with high correlations. The initial threshold setting should consider the actual conditions in the target area and the characteristics of the services. For example, you can determine an appropriate initial value based on historical data or expert experience.

[0081] Using a service item correlation analysis algorithm, we break down the service item feature data into tasks with clear boundaries. This algorithm calculates the correlation between each service item and other service items. When the correlation exceeds an initial threshold, these services are considered strongly connected and can be broken down into independent service units. This way, the entire service item feature data is broken down into N independent service units, each with relatively independent functions and tasks.

[0082] Label the process nodes for independent service units. This process node labeling involves marking and defining each operational step for each independent service unit, clearly specifying the start and end time, operation content, and required resources. This process node labeling provides a clear understanding of the execution flow of each independent service unit.

[0083] After completing the process node annotation, we construct an execution path diagram for each independent service unit. The execution path diagram graphically displays the sequence and logical relationships between the various process nodes in the independent service unit, intuitively presenting the execution process of the independent service unit.

[0084] Based on the execution path diagram, the operation frequency, coverage, and resource input for each independent service unit are calculated. Operation frequency refers to the number of times the independent service unit is executed within a certain period of time, coverage refers to the area covered by the independent service unit, and resource input includes human, material, and financial resources. By calculating these parameters, a task decomposition feature set is obtained, which contains the key characteristic information of each independent service unit.

[0085] The spatial location information of the task decomposition feature set is obtained based on the service item feature data. The spatial location information is used to determine the specific location of each task decomposition feature set in the target area, such as the specific coordinates or range of the operating area of ​​an independent service unit in the target area.

[0086] Based on spatial location information, the task decomposition feature set for each location is mapped to a region to construct a three-dimensional task decomposition feature map. Region mapping maps the task decomposition feature set to the spatial location of the target area, displaying the task decomposition features of each location in three-dimensional space. This three-dimensional task decomposition feature map provides a more intuitive and comprehensive representation of the distribution and characteristics of tasks within the target area.

[0087] The task weight for each location is determined based on the three-dimensional task decomposition feature map. The task weight represents the importance of each location's task within the entire target area and is calculated based on the parameters of the location's task decomposition feature set. For example, a location with high frequency of operations, wide coverage, and significant resource investment may have a relatively high task weight.

[0088] Based on the task weights at each location, a task decomposition heat map and task weight gradient field are constructed for the target area. The task decomposition heat map uses color to represent task weights, with darker colors indicating higher task weights and darker colors indicating lower task weights. The task weight gradient field displays the changing trend and gradient distribution of task weights in the target area.

[0089] The task decomposition heat map and task weight gradient field provide a clear understanding of the importance and distribution of tasks at various locations within the target area, thereby determining the task decomposition within the target area. This determination of the task decomposition provides an important foundation for subsequent sanitation service evaluation, enabling more targeted and accurate evaluation.

[0090] Example 2: When constructing an environmental sanitation service evaluation index system based on operation process data and determining a feedback data collection strategy based on the evaluation index system and task decomposition, the specific implementation method is as follows:

[0091] Perform multi-dimensional weighted analysis on operational process data. This data contains detailed information about each step in the sanitation operation, such as cleaning schedules, equipment usage, and staffing. Multi-dimensional weighted analysis analyzes this data from different perspectives to determine the importance of each dimension within the overall operational process. For example, analysis can be performed from dimensions such as operational efficiency, operational quality, and resource consumption. Each dimension contains multiple specific indicators, and the weight of each dimension is determined by assessing their importance.

[0092] The indicator feature tensor is constructed through standardization. Standardization converts indicator data of different dimensions and scales into a unified, standardized form for comparison and analysis. When constructing the indicator feature tensor, the standardized indicator data is organized according to a specific structure, forming a multidimensional tensor that comprehensively reflects the characteristics of the workflow data.

[0093] The service quality factors are identified based on the execution signals of the operation process data based on the fuzzy comprehensive evaluation method. The fuzzy comprehensive evaluation method is an evaluation method for dealing with fuzzy problems, which can combine qualitative evaluation and quantitative evaluation. When analyzing the execution signals, considering that service quality factors often have a certain degree of fuzziness, such as there is no clear boundary between "good" and "bad" cleanliness, this method can more accurately identify service quality factors and generate a target area evaluation index system that includes operation frequency, cleanliness, response time, etc. Operation frequency refers to the number of times sanitation operations are performed within a certain period of time, cleanliness refers to the cleanliness of the operation area, and response time refers to the speed of response to sudden problems or tasks.

[0094] Perform a dimension mapping operation on the task decomposition and the evaluation index system. The task decomposition describes in detail the decomposition of each task in the target area, including task type, distribution, and weight. The evaluation index system is a series of indicators constructed from the perspective of service quality. The dimension mapping operation corresponds and associates each dimension in the task decomposition with the indicators in the evaluation index system, so that the task decomposition can be reflected in the evaluation index system.

[0095] A multidimensional service quality evaluation model was constructed through dimensional mapping. This model integrates information from task decomposition and the evaluation index system to form a multidimensional and multi-level evaluation framework that can evaluate the quality of sanitation services from multiple perspectives.

[0096] The multidimensional service quality evaluation model is divided into M time intervals according to a preset time window. The size of the preset time window can be set according to actual needs, for example, to one day, one week, or one month. By dividing the time intervals into segments, service quality within different time periods can be analyzed and evaluated.

[0097] The service quality score information for each time interval is obtained based on the task decomposition. The task decomposition includes the execution status of each task in different time intervals. By comparing and evaluating these execution status with the indicators in the evaluation index system, the service quality score information for each time interval is obtained.

[0098] The DBSCAN clustering algorithm is used to cluster time intervals with similar rating trends in the multidimensional service quality evaluation model. The DBSCAN clustering algorithm is a density-based clustering algorithm that clusters points with high spatial density into a single cluster. The algorithm uses the service quality rating information for each time interval as a data point, calculates the similarity between them, and clusters time intervals with similar rating trends into a single cluster, generating the clustering results.

[0099] The clustering results are used to determine the service quality assessment priority for each time interval. The clustering results group the time intervals into different categories, with time intervals within each category showing similar scoring trends. Priority is determined based on the score and importance of each category. For example, a category with a lower score may require a higher assessment priority to facilitate timely identification and resolution of issues.

[0100] Obtain the corresponding spatiotemporal information for each time interval within the target area. This information includes the start and end times of the time interval, as well as the area covered by the work within that time interval. By obtaining this spatiotemporal information, you can determine the specific location and time range of each time interval within the target area.

[0101] The service evaluation importance of each location in the target area is assessed based on the evaluation priority and corresponding spatiotemporal information, resulting in an evaluation importance score for each location. Evaluation importance assessment comprehensively considers evaluation priority and spatiotemporal information to determine the importance of each location in the service evaluation. For example, a location with a high evaluation priority and frequent operations is likely to have a higher evaluation importance score.

[0102] The evaluation requirements for each location in the target area are determined based on the evaluation importance score of each location. This includes whether data collection should be performed and the required collection density. For locations with high evaluation importance scores, data collection is required, and the collection density may be high to provide a more comprehensive and accurate understanding of the service quality at that location. For locations with low evaluation importance scores, the collection density may be appropriately reduced or omitted.

[0103] Get feedback data collection equipment for different areas of service information acquisition completeness data. Information acquisition completeness data reflects the completeness of feedback data collection equipment in different areas, which is affected by factors such as regional environment and equipment performance.

[0104] The distance at which the feedback data collection device collects data at each location in the target area is determined based on the information completeness data and assessment requirements. The determination of data collection distance information should take into account both information completeness and assessment requirements. For example, in areas with low information completeness, the collection distance may need to be shortened to improve data integrity. In areas with high assessment requirements, the appropriate collection distance should also be determined based on actual conditions.

[0105] The collection point set for the feedback data collection device is determined based on the data collection distance information. The collection point set refers to the locations of specific data collection points within the target area. These points must be selected to meet the data collection distance requirements and be able to comprehensively and accurately reflect the service quality of the target area.

[0106] Determine the collection strategy for the feedback data collection device based on the collection point set. This strategy includes the device's movement path, collection interval, and collection method. It's developed based on the collection point set and actual needs. It aims to ensure the feedback data collection device can efficiently and accurately obtain service feedback data in the target area, providing reliable data support for subsequent service quality evaluation and analysis.

[0107] Example 3: When determining the collection strategy of the feedback data collection device according to the collection point set, the specific implementation method is as follows:

[0108] The spatiotemporal coordinates of each collection point are obtained, along with the initial position of the data collection device. This information includes the geographic location of the collection point and the corresponding time interval, such as the latitude and longitude of a point within the target area, and the time period within which data collection is required. The initial position of the data collection device is the specific location of the device at the time it begins the collection task, serving as the starting point for subsequent path planning.

[0109] Identify operational blind spots within the target area based on an evaluation index system. This system includes metrics for operational standardization and cleanliness compliance. By analyzing the distribution of these metrics within the target area, we can identify areas where effective coverage is difficult due to complex environments, insufficient equipment coverage, or human factors. These areas are known as operational blind spots. For example, narrow alleyways or densely greened areas may be inaccessible to equipment, creating operational blind spots.

[0110] Path planning is performed based on the initial position information and the spatiotemporal coordinates of each collection point using the ant colony algorithm. The ant colony algorithm is a heuristic algorithm that simulates the foraging behavior of ants and finds the optimal path through the accumulation and updating of pheromones. When applying this algorithm, the initial position is considered the starting point of the ant colony, and the spatiotemporal coordinates of each collection point are used as nodes to be visited. The algorithm considers factors such as the distance between nodes and time constraints, calculates possible paths from the initial position to each collection point, and gradually optimizes them through iterative updates of pheromones, ultimately outputting the shortest initial collection path for the data collection device. Furthermore, blind spots are used as restricted areas for path planning to ensure that the planned path does not enter these inoperable areas, preventing the equipment from entering blind spots, resulting in collection mission failure or inaccurate data.

[0111] As the feedback data collection device collects data along the shortest initial collection path, it acquires real-time data on the device's state changes. This data includes the device's movement speed, direction, and sensor operating status. This data can be used to analyze the density of traffic and the intensity of operational interference at the device's real-time collection location. For example, if the device's movement speed slows significantly and the sensor detects significant human activity, it can be determined that the current location has a high density of traffic. Conversely, if the sensor receives signals from other operating equipment or data such as noise and vibration in the environment, it can be used to analyze the intensity of operational interference.

[0112] Obtain feedback on the device's anti-interference capabilities, including its ability to filter out interference signals of varying intensities. For example, when facing high-intensity electromagnetic interference, the device can filter out enough interference signals to ensure data accuracy. Also, the device's sensor efficiency in filtering out interference from crowds in crowded areas can be measured. This data reflects the device's performance in various interference environments.

[0113] Obtain feedback data on the device's response time to crowd density and work disruption intensity. This response time is the interval between the device detecting a change in crowd density or work disruption intensity and responding accordingly. This response time data can be used to determine the device's acquisition path adjustment lag. This lag is caused by the device's time required for recognition and response, resulting in a lag between path adjustments and the actual occurrence of disruption. This lag is the adjustment lag.

[0114] Based on the anti-interference capability data and the acquisition path adjustment hysteresis, the human traffic density and operational interference intensity at the device's real-time acquisition location are analyzed to determine the cumulative acquisition path deviation within the device's recognition response time. Cumulative path deviation refers to the cumulative deviation of the device's actual travel path from the shortest initial acquisition path due to interference during the recognition response time. For example, if the device takes 0.5 seconds to respond after detecting interference, and during this 0.5 second period, the device may deviate from the original path by a certain distance due to human traffic or operational interference, the cumulative path deviation is calculated over multiple such time periods.

[0115] If the cumulative offset of the acquisition path is less than the preset value, it means that the interference has little impact on the device, and the device can basically collect according to the shortest initial acquisition path. At this time, the path offset direction and offset distance of the device for collection according to the shortest initial acquisition path are determined based on the cumulative offset of the acquisition path. For example, if the cumulative offset shows that the device has offset 1 meter to the east within a certain period of time, then the offset direction is east and the offset distance is 1 meter. Based on the offset direction and offset distance, the adjustment direction and adjustment distance of the device are determined to obtain adjustment data. For example, the adjustment direction is west and the adjustment distance is 1 meter to correct the offset of the device. Then, based on the adjustment data, the shortest initial acquisition path of the device during the real-time acquisition process is adjusted to obtain the first acquisition strategy. The first acquisition strategy is to fine-tune the original path in the case of a small offset to ensure that the device can return to the correct acquisition path as soon as possible.

[0116] If the accumulated offset of the collection path exceeds the preset value, it indicates that the interference has a significant impact on the device, preventing it from collecting data along the shortest initial collection path. Significant adjustments to the path are required. At this point, the device obtains data on the pedestrian density and operational interference intensity along the real-time collection path, and constructs an interference change map based on this data. The interference change map graphically displays the distribution and changes in interference along the real-time collection path, such as the pedestrian density at different locations and the intensity of operational interference. By analyzing the interference change map, the interference change trend in the service evaluation area within the target area can be determined, such as whether the interference is gradually increasing or showing periodic changes.

[0117] Interference trends are interpolated using spline interpolation, a mathematical method that fits a smooth curve based on known interference data points, thereby predicting interference information within a preset range of the shortest initial acquisition path. For example, if the interference intensity at several points along the path is known, spline interpolation can be used to estimate the interference intensity at other points along the path. Based on the interference information and interference mitigation data, the device's acquisition stability within the preset range of the shortest initial acquisition path is determined. Acquisition stability refers to whether the device can stably acquire data within this range without being significantly affected by interference. If acquisition stability is low, the path segment is unsuitable for data acquisition. Optimization is required for segments of the shortest initial acquisition path where the accumulated acquisition path deviation exceeds a preset value. Optimization methods include replanning the path to avoid high-interference areas and selecting routes with less interference, thereby obtaining an updated acquisition path. The feedback data acquisition device performs acquisition operations based on the updated acquisition path, generating a second acquisition strategy. This second acquisition strategy replans the path in the presence of severe interference to ensure that the device can efficiently and accurately complete data acquisition tasks in complex interference environments.

[0118] Example 4: When the feedback data collection device obtains service feedback data of the target area according to the collection strategy, identifies the service quality level and defect items to obtain service evaluation data, and transmits the service evaluation data to the management control center for indicator correction, the specific implementation method is as follows:

[0119] Obtain standard feedback data for different service items in the target area. This standard feedback data is based on industry norms, corporate standards, or historical experience data, and covers the ideal feedback status of various sanitation services in the target area. For example, for road sweeping services, standard feedback data may include post-sweeping road cleanliness indicators and garbage residue thresholds; for garbage collection and transportation services, it may include collection and transportation intervals and garbage container overflow rate standards.

[0120] Standard feedback data is labeled with a service level. This involves assigning a corresponding service quality level label to each set of standard feedback data based on pre-defined grading rules. For example, service quality can be categorized as excellent, good, acceptable, or unacceptable. The standard feedback data is then compared against each level's threshold to determine its level. This method generates labeled feedback data, providing clearly labeled training samples for subsequent model training.

[0121] A service quality identification model is constructed based on the random forest algorithm. The random forest algorithm is an ensemble learning method that improves model accuracy and stability by constructing multiple decision trees and combining their predictions. When building the model, the input features are determined to be the various indicators in the standard feedback data, and the output is the service quality level. Model parameter settings, such as the number of decision trees and the number of features considered when splitting each node, can be initialized based on actual needs and data characteristics.

[0122] Import the labeled feedback data into the service quality identification model for training. During training, the model automatically learns the mapping between various indicators in the standard feedback data and service quality levels. By continuously adjusting the model parameters, the model's prediction accuracy for the labeled feedback data reaches the desired level, resulting in a fully trained service quality identification model.

[0123] Feedback data collection equipment collects data within the target area according to a pre-determined collection strategy, acquiring service feedback data. This data represents feedback generated during actual operations and reflects the current state of sanitation services. For example, sensors installed on sanitation vehicles collect road cleanliness data, while GPS devices record operation times and routes.

[0124] The acquired service feedback data is fed into the trained service quality identification model. The model then identifies the service quality level based on the learned mapping relationships. The model outputs a service quality level for each service item, such as excellent or good. The model also compiles statistics on defective items for each service item, identifying indicators that deviate from the standard feedback data. These indicators are considered defective. For example, if the garbage collection interval exceeds the standard value, this indicator is identified as a defective item. This process yields service evaluation data, which includes the quality level and defective item information for each service item within the target area.

[0125] After obtaining service evaluation data, a layered data correction protocol stack is constructed based on a closed-loop feedback mechanism. This layered data correction protocol stack is a structured data processing and transmission framework that divides the data processing process into multiple layers, each responsible for a specific function, to ensure the accuracy and reliability of data transmission. For example, it may include a data acquisition layer, a feature extraction layer, a transmission layer, and a processing layer.

[0126] Feature extraction is performed on service evaluation data to construct a service evaluation feedback signal. Feature extraction extracts key information from service evaluation data that reflects its essential characteristics, such as the distribution characteristics of service quality levels and the types and frequencies of defects. Through feature extraction, the raw service evaluation data is converted into a service evaluation feedback signal suitable for transmission and processing.

[0127] The service evaluation feedback signals are transmitted to the management and control center using the layered data correction protocol stack. During transmission, each layer of the protocol stack performs data processing, such as encryption, compression, and error checking, to ensure data is not lost or tampered with during transmission. The management and control center is the core control unit of the entire sanitation service evaluation system, responsible for receiving and processing feedback signals from feedback data collection devices.

[0128] After receiving the service evaluation feedback signal, the management and control center performs an indicator mapping operation. This operation maps and correlates the characteristic information in the service evaluation feedback signal with the various indicators in the sanitation service evaluation index system, thereby obtaining corrected sanitation service evaluation data for the target area. For example, if the service evaluation feedback signal indicates frequent deficiencies in the cleanliness indicator, the indicator mapping operation will map this information to the cleanliness indicator in the evaluation index system, providing a basis for subsequent indicator corrections.

[0129] Example 5: When transmitting the service evaluation feedback signal to the management control center for index correction, the specific implementation method is as follows:

[0130] Obtain the management and control center's historical evaluation database. This database stores historical sanitation service evaluation data for the target area, including service evaluation indicators, service quality ratings, and defect records for different time periods. For example, the database might contain monthly road cleaning cleanliness data and garbage collection and transportation timeliness data for the past year, along with the corresponding evaluation results.

[0131] Similarity matching is performed on the service evaluation feedback signal against the historical evaluation database. Similarity matching is performed by calculating the feature differences between the service evaluation feedback signal and the historical evaluation data to determine the degree of similarity between them. For example, if the current service evaluation feedback signal shows that the on-time collection rate for a certain area is 75%, and the cleanliness defects are mainly caused by fruit peel residue, these features can be matched with records in the historical database where the on-time collection rate is between 70% and 80%, and the defects include fruit peel residue. This matching result reflects the degree of similarity and correspondence between the current feedback signal and the historical data.

[0132] The matching results are used to determine the distribution characteristics of abnormal data in the service evaluation feedback signal. Abnormal data refers to data that differs significantly from historical data and falls outside the normal fluctuation range. For example, if the average road sweeping cleanliness in historical data is 90%, but the cleanliness of a certain area in the current feedback signal is only 60%, and the matching results show that there has been no special construction in the area under similar historical circumstances, then this data can be judged as abnormal. By analyzing the matching results, the temporal and spatial distribution of abnormal data can be determined, such as which time periods and areas have the most abnormal data.

[0133] A sliding window algorithm is used to perform time-series filtering on abnormal data. This algorithm sets a fixed-length time window, slides it across the time series data, and analyzes and processes the data within the window. For example, a seven-day window can be set, sliding forward from the time of the current feedback signal. Abnormal data within each window is then counted and analyzed, filtering out abnormal data caused by short-term fluctuations or accidental factors while retaining abnormal data with persistence or regularity. This results in valid feedback data. Valid feedback data can more realistically reflect existing problems in sanitation services.

[0134] Principal component analysis is performed on valid feedback data to extract key factors influencing service quality. Principal component analysis is a dimensionality reduction method that extracts a small number of composite variables from multiple variables. These composite variables can reflect the key information of the original variables. For example, valid feedback data may include multiple indicators such as cleanliness, operation frequency, and response time. Through principal component analysis, key factors such as "operation execution efficiency" and "environmental maintenance effectiveness" can be extracted. These key factors can generally reflect the main factors affecting service quality.

[0135] Determine the corresponding relationship between key factors and the evaluation index system. The evaluation index system includes indicators for multiple dimensions, such as operational standardization and cleanliness compliance rate. Key factors need to be associated with these indicators to clearly define the specific indicators affected by each key factor. For example, the "environmental maintenance effect" key factor may be associated with indicators such as road cleaning cleanliness and garbage collection and transportation timeliness under the cleanliness compliance rate dimension.

[0136] Based on the covariance matrix, the correlation coefficients between each evaluation indicator and key factors are calculated to construct a modified indicator correlation matrix. The covariance matrix is ​​used to measure the correlation between variables. The correlation coefficients are obtained by calculating the covariances between each evaluation indicator and key factors and performing normalization. The larger the absolute value of the correlation coefficient, the stronger the correlation between the indicator and the key factor. These correlation coefficients are arranged in a specific structure to construct a modified indicator correlation matrix, which clearly shows the degree of correlation between each evaluation indicator and key factor.

[0137] Sort the correlation coefficients of each indicator in the correlation matrix and select evaluation indicators whose absolute values ​​of the correlation coefficients are less than the preset threshold as those requiring adjustment. The preset threshold can be set based on actual needs and historical experience, for example, 0.3. Indicators with absolute values ​​of correlation coefficients less than the threshold indicate a weak correlation with the key factor and may not accurately reflect service quality in the current service evaluation, requiring adjustment. For example, if the correlation coefficient of the "Equipment Usage Compliance" indicator with the "Job Execution Efficiency" key factor is 0.25, which is less than the threshold of 0.3, then this indicator will be selected as an indicator requiring adjustment.

[0138] For quantitative indicators that require adjustment, the upper and lower thresholds will be reset based on the statistical distribution of valid feedback data. Quantitative indicators are those that can be measured using specific numerical values, such as operation frequency and cleanliness values. For example, the original threshold for on-time garbage collection and transportation rates was set at an upper limit of 95% and a lower limit of 80%. If valid feedback data indicates that most on-time collection and transportation rates are between 85% and 90%, and historical data indicates good service quality within this range, the thresholds can be adjusted to an upper limit of 90% and a lower limit of 75% to better reflect actual operational conditions.

[0139] For qualitative indicators in the evaluation index items that need to be adjusted, the level description will be redefined based on the semantic analysis results of the effective feedback data. Qualitative indicators are indicators measured by text descriptions, such as the levels of "satisfied", "average", and "unsatisfied" for complaint handling satisfaction. For example, the description of the "good" level in the original "cleanliness" qualitative indicator is "no obvious garbage on the road surface", while the semantic analysis of the effective feedback data shows that users have different understandings of "no obvious garbage". Some users believe that fallen leaves are "obvious garbage". The description of the "good" level can be redefined as "no garbage in pieces or waste with a diameter greater than 5 cm on the road surface, and the amount of fallen leaves does not exceed 10 pieces per square meter" to make the level description clearer and more specific.

[0140] The weight coefficients of all evaluation indicators are normalized and adjusted to ensure that the total weight is 1, generating an updated environmental sanitation service evaluation indicator system. The weight coefficient reflects the importance of each indicator in the evaluation system. After adjusting some indicators, the weights of all indicators need to be redistributed and normalized. For example, if the weight of a certain indicator under the "cleanliness compliance rate" dimension is reduced, the weights of other more relevant indicators can be appropriately increased accordingly, ultimately ensuring that the sum of all indicator weights is 1. The updated evaluation indicator system can more accurately reflect the current actual situation of environmental sanitation services and improve the scientific and rational nature of the evaluation.

[0141] When constructing a multidimensional service quality evaluation model, we first identified the core dimensions of service quality assessment, including operational standardization, cleanliness compliance, timely response, and complaint handling rate. Specific evaluation indicators were set for each core dimension. For example, operational standardization encompasses equipment compliance and operational process integrity; cleanliness compliance encompasses road sweeping cleanliness and timely garbage collection and transportation; timely response encompasses emergency response time and task completion time; and complaint handling rate encompasses complaint acceptance rate and satisfaction with problem resolution.

[0142] Experts were organized to rate the importance of each dimension and indicator using the Delphi method. The Delphi method involves multiple rounds of anonymous expert consultation to gradually reach consensus. For example, 10 sanitation industry experts were invited to rate the importance of each dimension and indicator (1-10). The scores were collected and statistically analyzed, with the highest and lowest scores removed and the average score calculated. Based on the average score, the weight coefficient for each dimension and the sub-weight coefficient for each indicator within the corresponding dimension were calculated.

[0143] By integrating the evaluation indicators, weight coefficients, and sub-weight coefficients of each dimension, a multidimensional service quality evaluation model with multi-dimensional indicators and hierarchical weights was constructed. For example, the weight of the operational standardization dimension was 30%, of which the sub-weight of equipment compliance was 60%, and the sub-weight of operational process integrity was 40%. The weight of the cleanliness compliance dimension was 40%, of which the sub-weight of road sweeping cleanliness was 55%, and the sub-weight of garbage collection and transportation timeliness was 45%. Through this structured integration, a complete, multi-layered service quality evaluation model was formed, providing a scientific assessment framework for sanitation service evaluation.

[0144] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0145] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism, characterized in that: The following steps are involved: Acquire sanitation operation process data and service item characteristic data of the target area based on the task decoupling model, and determine the task decomposition of the target area according to the service item characteristic data; Constructing an environmental sanitation service evaluation index system based on the operation process data, and determining a feedback data collection strategy based on the evaluation index system and task decomposition; The feedback data collection device obtains service feedback data of the target area according to the collection strategy, identifies the service quality level and defect items of the target area according to the feedback data, and obtains service evaluation data; Transmitting the service evaluation data to the management control center for index correction according to a closed-loop feedback mechanism; The task decoupling model is used to obtain the sanitation operation process data and service item characteristic data of the target area, and the task decomposition of the target area is determined according to the service item characteristic data, specifically: Acquire the operation process data of the target area and the characteristic data of the service items in the target area based on the task decoupling model; Adaptively classifying the service item feature data based on a workflow decomposition algorithm to obtain service item feature data with clear boundaries, introducing a service item relevance analysis algorithm, and setting initial thresholds for parameters of the service item relevance analysis algorithm; Decomposing the feature data of the service items with clear boundaries into tasks according to the service item correlation analysis algorithm to obtain N independent service units; Marking the process nodes of the independent service units, constructing an execution path graph for each independent service unit, and calculating the operation frequency, coverage, and resource input of each independent service unit based on the execution path graph to obtain a task decomposition feature set; Acquire spatial location information of the task decomposition feature set according to the service item feature data, perform regional mapping of the task decomposition feature set at each location according to the spatial location information, and construct a three-dimensional task decomposition feature map; The task weight of each position is determined according to the three-dimensional task decomposition feature map, the task decomposition heat map and task weight gradient field of the target area are constructed according to the task weight of each position, and the task decomposition status of the target area is determined according to the task decomposition heat map and task weight gradient field.

2. The sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism according to claim 1 is characterized in that: The sanitation service evaluation index system is constructed based on the operation process data, and the feedback data collection strategy is determined based on the evaluation index system and task decomposition, specifically: Perform multi-dimensional weight analysis on the operation process data, construct an indicator feature tensor through standardization processing, identify service quality factors of the execution signal of the operation process data based on the fuzzy comprehensive evaluation method, and generate an evaluation index system for target areas including operation frequency, cleanliness, and response time; Performing a dimension mapping operation on the task decomposition and the evaluation index system to construct a multidimensional service quality evaluation model, and dividing the multidimensional service quality evaluation model according to a preset time window to construct M time intervals; Obtaining service quality score information for each time interval according to the task decomposition, and clustering time intervals with similar score trends in the service quality multidimensional evaluation model based on the DBSCAN clustering algorithm to obtain clustering results; Determining an evaluation priority of the service quality of each time interval according to the clustering result, obtaining corresponding spatiotemporal information of each time interval in the target area, and performing a service evaluation importance evaluation on each location in the target area according to the evaluation priority and the corresponding spatiotemporal information to obtain an evaluation importance score for each location in the target area; Determining evaluation requirement information for each location in the target area according to the evaluation importance score of each location, wherein the evaluation requirement information includes whether to perform collection and collection density requirement information; Obtaining information acquisition completeness data of the feedback data acquisition device for different areas of service, determining data collection distance information of the feedback data acquisition device for each location in the target area based on the information acquisition completeness data and the evaluation requirement information, and determining a collection point set of the feedback data acquisition device based on the data collection distance information; A collection strategy of the feedback data collection device is determined according to the collection point set.

3. The sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism according to claim 2 is characterized in that: The collection strategy of the feedback data collection device is determined according to the collection point set, specifically: Obtaining the spatiotemporal coordinate information of each collection point and the initial position information of the feedback data collection device, and determining the operation blind area in the target area according to the evaluation index system; Performing path planning on the initial position information and the spatiotemporal coordinate information of each collection point based on an ant colony algorithm, taking the operation blind area as a path planning restriction area, and outputting the shortest initial collection path for the feedback data collection device; Acquire in real time device state change data of the feedback data acquisition device during the acquisition process according to the shortest initial acquisition path, and determine the crowd density and operation interference intensity at the real-time acquisition location of the feedback data acquisition device based on the state change data; Acquiring anti-interference capability data of the feedback data acquisition device, wherein the anti-interference capability data includes information filtering capability data of the feedback data acquisition device for interferences of different intensities; Obtaining recognition response time data of the feedback data acquisition device to the crowd density and the work interference intensity, and determining a collection path adjustment hysteresis amount of the feedback data acquisition device according to the recognition response time data; Analyzing the crowd density and the intensity of the work interference at the real-time collection location of the feedback data collection device according to the anti-interference capability data and the collection path adjustment hysteresis, and determining the cumulative amount of collection path deviation of the feedback data collection device within the recognition response time; If the accumulated amount of the acquisition path offset is less than a preset value, determining a path offset direction and an offset distance for the feedback data acquisition device to acquire data according to the shortest initial acquisition path based on the accumulated amount of the acquisition path offset, and determining an adjustment direction and an adjustment distance of the feedback data acquisition device based on the offset direction and the offset distance to obtain adjustment data; Adjusting the shortest initial acquisition path of the feedback data acquisition device during the real-time acquisition process according to the adjustment data to obtain a first acquisition strategy; If the accumulated amount of the acquisition path deviation is greater than a preset value, obtaining the pedestrian flow density and operation interference intensity data of the real-time acquisition path of the feedback data acquisition device, constructing an interference change map based on the pedestrian flow density and operation interference intensity data of the real-time acquisition path, and determining the interference change trend of the service evaluation area in the target area based on the interference change map; An interpolation operation is performed on the interference change trend based on spline interpolation to determine the interference information within the preset range of the shortest initial acquisition path. The acquisition stability of the feedback data acquisition device within the preset range of the shortest initial acquisition path is determined based on the interference information and the anti-interference capability data. According to the acquisition stability, the shortest initial acquisition path section with the accumulated acquisition path offset greater than the preset value is optimized to obtain an updated acquisition path. The feedback data acquisition device performs an acquisition operation based on the updated acquisition path to obtain a second acquisition strategy.

4. The sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism according to claim 1 is characterized in that: The feedback data collection device obtains service feedback data of the target area according to the collection strategy, identifies the service quality level and defect items of the target area according to the feedback data, and obtains service evaluation data, specifically: Obtaining standard feedback data for different service items in the target area, and labeling the standard feedback data with service levels to obtain labeled feedback data; Building a service quality recognition model based on a random forest algorithm, and importing the labeled feedback data into the service quality recognition model for training; The feedback data collection device obtains service feedback data of the target area according to the collection strategy, imports the service feedback data into the trained service quality identification model to identify the service quality level, and counts the defective items of each service item to obtain service evaluation data.

5. The sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism according to claim 1 is characterized in that: The service evaluation data is transmitted to the management control center according to the closed-loop feedback mechanism for index correction, specifically: Building a layered data correction protocol stack based on a closed-loop feedback mechanism, extracting features from the service evaluation data, and building a service evaluation feedback signal; The service evaluation feedback signal is transmitted to a management control center according to the layered data correction protocol stack, and an indicator mapping operation is performed on the service evaluation feedback signal to obtain the sanitation service evaluation correction data of the target area.

6. The sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism according to claim 5 is characterized in that: The service evaluation feedback signal is transmitted to the management control center according to the layered data correction protocol stack, and the service evaluation feedback signal is subjected to an indicator mapping operation to obtain the sanitation service evaluation correction data of the target area, specifically: Obtaining a historical evaluation database of a management control center, performing similarity matching between the service evaluation feedback signal and the historical evaluation database to obtain a matching result; Determine the distribution characteristics of abnormal data in the service evaluation feedback signal based on the matching results, and perform a time series filtering operation on the abnormal data based on a sliding window algorithm to obtain valid feedback data; Aligning the effective feedback data with the evaluation index system, constructing an index correction correlation matrix, and determining the evaluation index items that need to be adjusted according to the weight change trend of each index in the correlation matrix; The threshold range or weight coefficient of the evaluation index items that need to be adjusted is corrected to generate an updated sanitation service evaluation index system, and the updated evaluation index system is stored in the management and control center to obtain the sanitation service evaluation correction data of the target area.

7. The sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism according to claim 6 is characterized in that: The effective feedback data is dimensionally aligned with the evaluation index system to construct an index correction correlation matrix, and the evaluation index items that need to be adjusted are determined according to the weight change trend of each index in the correlation matrix, specifically: Performing principal component analysis on the effective feedback data to extract key factors affecting service quality, and determining corresponding correlation relationships with the evaluation index system based on the key factors; Calculate the correlation coefficients between each evaluation index and key factors based on the covariance matrix and construct the index correction correlation matrix; The correlation coefficients of the indicators in the correlation matrix are sorted, and evaluation indicator items whose absolute values ​​of the correlation coefficients are less than a preset threshold are screened out as evaluation indicator items that need to be adjusted.

8. The sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism according to claim 7 is characterized in that: The threshold range or weight coefficient of the evaluation index item that needs to be adjusted is modified to generate an updated environmental sanitation service evaluation index system, specifically: For quantitative indicators in the evaluation indicators that need to be adjusted, the upper and lower thresholds will be reset based on the statistical distribution of valid feedback data; For qualitative indicators in the evaluation indicators that need to be adjusted, redefine the level description based on the semantic analysis results of the effective feedback data; Normalize and adjust the weight coefficients of all evaluation index items to ensure that the total weight is 1, and generate an updated sanitation service evaluation index system.

9. The sanitation service evaluation method based on task decoupling and closed-loop feedback mechanism according to claim 2 is characterized in that: The construction of the multi-dimensional service quality evaluation model is specifically as follows: Determine the core dimensions of service quality assessment, including operational standardization, cleanliness compliance rate, response timeliness, and complaint handling rate; Specific evaluation indicators are set for each core dimension, including: operational standardization includes equipment compliance and operational process integrity; cleaning compliance rate includes road sweeping cleanliness and garbage collection and transportation timeliness; response timeliness includes emergency response time and task dispatch completion timeliness; complaint handling rate includes complaint acceptance rate and problem resolution satisfaction; Based on the Delphi method, an expert algorithm is used to evaluate the importance of each dimension and indicator. The weight coefficient of each dimension and the sub-weight coefficient of each indicator within the corresponding dimension are calculated according to the scoring results. The evaluation indicators, weight coefficients and sub-weight coefficients of each dimension are structurally integrated to construct a multidimensional evaluation model of service quality that includes multidimensional indicators and hierarchical weights.

Images