Urban steward service method and system based on Internet of Things

Abstract

The invention provides a city steward service method and system based on the Internet of Things, and relates to the technical field of city steward. Comprising the following steps: dividing a management unit according to urban area features, defining a facility sensing template based on functions of the management unit, and classifying sensing equipment according to monitoring types to obtain equipment sensing data; based on a federated learning algorithm, fusing the multi-source sensing data to construct a city state evaluation model, inputting equipment sensing data and historical operation and maintenance records, and generating a city operation state evaluation result; according to a city operation state evaluation result, the city management demand is disassembled into a plurality of service tasks, and equipment perception data is associated to construct a service response matrix; and calculating a priority coefficient of each service task in the service response matrix, and dynamically scheduling resources to generate an optimal service scheme. According to the method, the city state evaluation model is constructed by fusing multi-source sensing data through the federated learning algorithm, the dynamic change of the city is continuously adapted, and the city management is promoted to be more intelligent.

Description

Technical Field

[0001] This invention relates to the field of urban concierge technology, and in particular to an urban concierge service method and an urban concierge service system based on the Internet of Things. Background Technology

[0002] Urban management is a crucial aspect of ensuring the normal operation of cities and improving the quality of life for residents. With the acceleration of urbanization, the scale of cities continues to expand, and the population continues to grow, urban management faces numerous challenges, such as untimely maintenance of public facilities, difficulties in environmental sanitation supervision, chaotic traffic order, and delayed response to emergency events.

[0003] Traditional urban management methods rely heavily on manual inspections and feedback, which suffers from inefficiency, limited coverage, and untimely information transmission. When public facilities malfunction, repairs are often only arranged after residents report them or after inspection personnel discover them, which can lead to prolonged delays in addressing the problems and disrupting residents' lives. In terms of environmental sanitation, manual inspections struggle to monitor the cleanliness of each area in real time, easily resulting in garbage accumulation. In the face of emergencies, delayed information access and poor inter-departmental coordination can delay optimal response times, leading to resource waste and hindering the efficient and orderly operation of urban services. Summary of the Invention

[0004] This invention provides a method and system for urban concierge services based on the Internet of Things, in order to solve the defects of low data utilization and resource waste in the existing technology.

[0005] On one hand, the present invention provides a method for providing urban concierge services based on the Internet of Things, including: S1: Divide the city into management units based on the characteristics of the urban area, deploy IoT sensing devices to collect urban element data, define facility sensing templates based on the functions of the management units, and classify the sensing devices according to the monitoring type to obtain device sensing data.

[0006] S2: Based on the federated learning algorithm, a city status assessment model is constructed by fusing multi-source sensing data. Input device sensing data and historical operation and maintenance records to generate city operation status assessment results.

[0007] S3: Based on the city's operational status assessment results, service standards, and resource allocation, the city's management needs are broken down into multiple service tasks, and a service response matrix is ​​constructed by linking the sensing data of related devices with operation and maintenance resources.

[0008] S4: Calculate the priority coefficient of each service task in the service response matrix and dynamically schedule resources to generate the optimal service solution.

[0009] According to the urban concierge service method based on the Internet of Things provided by the present invention, the specific steps for obtaining device sensing data in step S1 are as follows: S11: Divide the management units according to urban functional areas and geographical boundaries, label the core facility type in each unit, and determine the perception indicators based on the characteristics of the facilities within the unit.

[0010] S12: Extract features from management units based on core facility types and perception indicators to obtain core service requirements, set perception dimensions for each facility, define application scenarios based on core service requirements, and obtain facility perception templates.

[0011] S13: Collect raw data from sensing devices in the facility sensing template, and use image recognition technology to transform unstructured data in the raw data into structured raw data corresponding to the sensing indicators.

[0012] S14: Classify the structured raw data according to the detection type to obtain device sensing data.

[0013] According to the urban concierge service method based on the Internet of Things provided by the present invention, the specific steps for classifying the structured raw data in step S14 are as follows: The perception dimensions corresponding to the structured indicators are extracted from the facility perception template.

[0014] The weights of the perception dimensions are calculated based on the importance of the facilities in the management unit, and the structured indicators are compared with the facility perception template to obtain the equipment data quality score.

[0015] The overall effectiveness score of the device data is calculated based on the weight of the perception dimension and the device data quality score. The structured raw data is then divided according to the score to obtain the device perception data.

[0016] According to the urban steward service method based on the Internet of Things provided by the present invention, the specific steps for obtaining the urban operation status assessment result in step S2 are as follows: S21: Clean the equipment sensing data and historical operation and maintenance records, remove outliers, supplement missing values ​​using time series interpolation, and extract time and spatial features to obtain preprocessed sensing data.

[0017] S22: Construct a federated learning framework based on preprocessed perceptual data, using deep neural networks as the local model for federated learning.

[0018] S23: Local nodes use their own device perception data and historical operation and maintenance records to train local models, calculate the update gradient of local model parameters, and upload the encrypted gradient information to the federated learning server to build a city status assessment model.

[0019] S24: The federated learning server uses a secure aggregation algorithm to aggregate and calculate the encrypted gradient information, obtain the updated values ​​of the parameters of the city status assessment model, and update the city status assessment model.

[0020] S25: Input the new equipment perception data and historical operation and maintenance records into the trained urban status assessment model, and output the urban operation status assessment results through model inference.

[0021] According to the urban concierge service method based on the Internet of Things provided by the present invention, the specific steps of training the local model by the local node in step S23 are as follows: Initialize the parameters of the deep neural network model.

[0022] The preprocessed sensing data is input into the local model, and forward propagation calculation is performed using the gradient descent method to obtain the prediction result.

[0023] The loss value is calculated using mean squared error based on the difference between the predicted results and the true labels.

[0024] The gradient corresponding to the local parameters is calculated based on the loss value using the backpropagation algorithm.

[0025] Based on the gradient, stochastic gradient descent is used to update the parameters of the local model, completing one local model training iteration. This process is repeated multiple times until the preset training stopping condition is met.

[0026] According to the urban concierge service method based on the Internet of Things provided by the present invention, the specific steps for constructing the service response matrix in step S3 are as follows: S31: Extract core management needs based on the city status assessment results to obtain a list of core management needs.

[0027] S32: Based on the currently available operation and maintenance resources and the service types in the core management requirements list, establish the correspondence between resource attributes and service capabilities.

[0028] S33: Based on the service type, the core requirements are broken down into several interrelated sub-tasks according to the service process and professional division of labor.

[0029] S34: Establish an initial matrix using subtasks and resources, adjust the initial matrix using data adaptation and priority, and obtain the service response matrix.

[0030] According to the urban concierge service method based on the Internet of Things provided by the present invention, the specific steps for adjusting the service response matrix in step S34 are as follows: S341: Construct an initial matrix with subtasks as rows and operation and maintenance resources as columns. Evaluate the adaptability based on adaptability and priority to obtain the initial matrix.

[0031] S342: Update and correct resource parameters and adjust the initial matrix according to service standards and resource configuration.

[0032] According to the urban concierge service method based on the Internet of Things provided by the present invention, the specific steps for calculating the priority coefficient of each service task in the service response matrix in step S4 are as follows: Based on the attribute information of each service task in the service response matrix, the core parameters affecting priority are extracted, a weight value is assigned to each parameter, the priority coefficient of each service task is calculated by weighted summation, and all service tasks are sorted according to the priority coefficient.

[0033] According to the urban concierge service method based on the Internet of Things provided by the present invention, the specific steps in step S4 of dynamically scheduling resources based on priority coefficients and generating an optimal service plan are as follows: Based on the resource and task compatibility and real-time resource status in the service response matrix, the tasks are sorted by priority to match the optimal resources. Resources with high compatibility and low current load are allocated to high-priority tasks first. At the same time, the geographical distribution of resources is considered to shorten the response time. The optimal service plan is generated based on the resource allocation results.

[0034] On the other hand, the present invention also provides an Internet of Things-based urban concierge service system, comprising: The sensing data acquisition module is used to divide management units according to urban area characteristics, deploy IoT sensing devices to collect urban element data, define facility sensing templates based on management unit functions, and classify sensing devices according to monitoring type to obtain device sensing data.

[0035] The city operation status assessment module uses a federated learning algorithm to integrate multi-source sensing data to build a city status assessment model. It takes device sensing data and historical operation and maintenance records as inputs and generates city operation status assessment results.

[0036] The matrix construction module breaks down urban management needs into multiple service tasks based on urban operation status assessment results, service standards, and resource allocation, and constructs a service response matrix by associating device perception data with operation and maintenance resources.

[0037] The solution generation module calculates the priority coefficient of each service task in the service response matrix and dynamically schedules resources to generate the optimal service solution.

[0038] This invention provides an Internet of Things-based urban management service method that uses a federated learning algorithm to fuse multi-source sensing data to construct an urban status assessment model. This solves the problem of low data utilization and difficulty in achieving refined and intelligent urban management.

[0039] By dividing management units according to the characteristics of urban areas, clarifying the core facility types of each unit and determining the sensing indicators, IoT sensing devices can be deployed in a targeted manner to collect data. Furthermore, technologies such as image recognition can be used to transform the data into a structured form and further classify it into key, routine, and auxiliary data categories. This makes the data used for subsequent decision-making more accurate and targeted, avoids interference from chaotic data, and lays a solid foundation for accurately grasping the conditions of various aspects of the city.

[0040] By leveraging federated learning algorithms to integrate multi-source sensing data to construct an urban status assessment model, and fully integrating scattered information resources, this model can not only comprehensively consider the multi-dimensional situation of each management unit, but also output key indicators such as facility health and the urgency of service needs. This allows managers to clearly understand the city's operational status from a macro perspective, so as to make practical and holistic planning and control measures.

[0041] Based on the city status assessment results, management needs are broken down into multiple service tasks, and a service response matrix is ​​constructed by linking device perception data and operation and maintenance resources. This enables the rational allocation of resources according to the actual situation, prioritizing important and urgent tasks, avoiding resource waste, improving resource utilization efficiency, and maximizing the value of limited human, material, and equipment resources to ensure the efficient and orderly operation of city services.

[0042] The system calculates the priority coefficients of service tasks and dynamically schedules resources accordingly to generate optimal service plans, fully considering key factors such as the urgency level, scope of impact, and importance of facilities. This allows for the orderly arrangement of task execution, prioritizing important and urgent matters. Furthermore, when resources are limited and conflicts exist, it rationally allocates resources, reduces internal friction, shortens response time, improves the responsiveness and effectiveness of urban management, and continuously optimizes service plans, making urban management work more organized and scientific.

[0043] The entire service approach revolves around data collection, processing, analysis, and application. From data acquisition by sensing devices to using data to assess city status, break down tasks, and allocate resources, every step is data-driven. This enables city managers to make scientific decisions based on objective and accurate data. Furthermore, as the sensing data is updated in real time, the entire system can be continuously adjusted and optimized to adapt to the dynamic changes in the city, promoting smarter and more efficient urban management and effectively improving the quality of life for city residents and the city's sustainable development capabilities. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0045] Figure 1 This is a flowchart illustrating an Internet of Things-based urban concierge service method provided in an embodiment of the present invention; Figure 2 This is a module diagram of an Internet of Things-based urban concierge service system provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the process for constructing a service response matrix provided in an embodiment of the present invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0047] The following is combined Figures 1-3 This invention describes a method and system for providing urban concierge services based on the Internet of Things (IoT).

[0048] Figure 1 This is a flowchart illustrating an Internet of Things-based urban concierge service method provided in an embodiment of the present invention.

[0049] like Figure 1 As shown in the figure, an embodiment of the present invention provides a city butler service method based on the Internet of Things, the method comprising: S1: Divide the city into management units based on the characteristics of the urban area, deploy IoT sensing devices to collect urban element data, define facility sensing templates based on the functions of the management units, and classify the sensing devices according to the monitoring type to obtain device sensing data.

[0050] S11: Management units are divided according to urban functional areas and geographical boundaries. In commercial areas, the need for waste sorting is more complex due to high population mobility and diverse types of waste generated. Residential areas generate a relatively stable amount of household waste, but this directly impacts the quality of the living environment. Industrial parks may generate special industrial waste. Based on these different functional areas and geographical boundaries such as streets and communities, detailed divisions are made, with each unit labeled with its core facility type. Perception indicators are determined for the characteristics of the facilities within each unit. Specifically, perception indicators include the overflow level of different types of waste bins and the accuracy rate of residents' correct waste sorting. For waste transfer stations, after clarifying the core facility type, perception indicators include daily waste transfer volume, odor concentration within the station, and the operational status of waste compression equipment. Regarding facilities related to the development of waste harmless treatment technologies, temperature and humidity within the treatment workshop, waste incineration temperature, and chemical indicators of landfill leachate treatment are important perception indicators.

[0051] S12: Feature extraction is performed on the management units to obtain core service requirements. Combined with the facility lifecycle of the waste transfer station, the facilities in the full facility list correspond to the core facility types in the basic information table. In specific implementation, the core service requirement of the waste transfer station in the initial construction phase is reasonable planning and layout to meet the waste transfer volume of the surrounding area. During the operation phase, the focus is on efficient transfer and stable equipment operation. During upgrades and renovations, the focus is on improving processing capacity. Combining these different stages and the collaborative management requirements with front-end waste sorting and back-end harmless treatment, a full facility list covering all aspects of the waste transfer station is generated, ensuring that the facilities in the list accurately correspond to the core facility types set in the basic information table, such as transfer equipment and storage areas. The system sets perception dimensions for each facility based on monitoring dimensions (monitoring the amount of waste entering and leaving the waste transfer station, and monitoring the proportion of various types of waste at waste sorting and disposal points), data transmission frequency, early warning thresholds (earning an alert when the amount of waste accumulated at the waste transfer station reaches 80% of the design capacity, and an alert when the misclassification rate at waste sorting and disposal points exceeds a certain proportion), and operation and maintenance interfaces (connecting to corresponding operation and maintenance operation points such as waste transfer vehicle scheduling, equipment maintenance, and allocation of sorting guidance personnel). Application scenarios are defined according to core service requirements, and facility perception templates are generated by combining the perception dimensions.

[0052] S13: Collect raw data from sensing devices in the facility sensing template, and use image recognition technology to transform unstructured data in the raw data into structured raw data corresponding to sensing indicators. At waste sorting and disposal points, images of residents disposing of waste are captured by cameras as raw data. These images are unstructured data. Image recognition technology is used to analyze the types of waste disposed of by residents in the images, whether they are disposed of correctly, etc., and transform them into statistically significant structured raw data such as "correct disposal rate of recyclable waste" and "quantity of incorrectly disposed hazardous waste". For waste transfer stations, images captured by environmental monitoring cameras within the station are subjected to image recognition to analyze whether waste is piled up in a standardized manner, whether there are any abnormal objects around the equipment, etc., and transform them into corresponding structured data. This facilitates a more accurate understanding of the actual situation within the station. At the same time, images of relevant operational processes and equipment operation appearance in the waste harmless treatment workshop are also identified and processed to extract key structured data for measuring whether the treatment process is compliant and whether the equipment is operating normally.

[0053] S14: The perception dimensions corresponding to the structured indicators are extracted from the facility perception template to evaluate the validity of equipment data. The weights of the perception dimensions are determined based on the importance of the facility within the management unit. The structured indicators are compared with the template to obtain a data quality score. The overall validity score of the equipment data is calculated by combining the weights and scores, and the data is categorized into key data, routine data, and auxiliary data according to the score. Specifically, when evaluating the validity of equipment operation data at a waste transfer station, perception dimensions such as real-time temperature and pressure are given higher weights because they are closely related to whether the equipment is operating normally. The weight of equipment appearance image data is relatively lower. By comparing with the preset facility perception template, if the actually collected data is accurate and complete in the key perception dimensions, the data quality score is high. A comprehensive validity score is then calculated to distinguish data types of different importance.

[0054] S2: Based on the federated learning algorithm, a city status assessment model is constructed by fusing multi-source sensing data. Input device sensing data and historical operation and maintenance records to generate city operation status assessment results.

[0055] S21: Preprocess the multi-source equipment sensing data and historical operation and maintenance records, including cleaning outliers (such as abnormally high or low waste volume data at waste transfer stations due to temporary sensor malfunctions, and incorrect classification data at waste sorting points due to obstruction, etc.), filling in missing values ​​(if key process parameter records are missing in the waste harmless treatment workshop for a certain period, they should be reasonably filled in based on data from previous and subsequent periods and related patterns), standardizing the data format, and dividing the data into multiple local nodes according to the data source to obtain preprocessed sensing data. Data from waste sorting points in different areas can be set as one local node, data from each waste transfer station can be set as an independent local node, and data from different waste harmless treatment facilities can also be set as corresponding local nodes.

[0056] S22: Based on the preprocessed multi-source data, a federated learning framework is constructed, the model parameter initialization scheme for each local node is determined, and a deep neural network is used as the local model for federated learning. For local nodes related to the operation and management of waste sorting projects, a neural network structure suitable for analyzing residents' sorting behavior patterns, sorting accuracy trends, etc., can be constructed.

[0057] S23: Each local node uses its own device perception data and historical operation and maintenance records to train its local model, calculates the update gradient of the model parameters, and uploads the encrypted gradient information to the federated learning server without disclosing the original data. The updated city state assessment model parameters are distributed to each local node, and each node continues to train its local model based on the new parameters, repeating the parameter update and aggregation process until the model converges.

[0058] The specific steps for training a local model based on local nodes are as follows: Initialize the parameters of the deep neural network model. In the neural network related to waste sorting and disposal points, the number of neurons in the input layer is determined based on the number of perceived indicators, the number of layers and neurons in the intermediate hidden layers are initialized according to the data complexity and model complexity requirements, and the output layer is set according to the indicators to be predicted, such as the classification accuracy and residents' classification behavior trends.

[0059] The preprocessed sensing data is input into the local model, and forward propagation calculations are performed using the gradient descent method to obtain the prediction results. Preprocessed data from waste sorting and disposal points, waste transfer stations, and waste disposal sites are input into their respective local models, and forward propagation calculations are performed using the gradient descent method. The waste sorting and disposal point model yields predicted values ​​such as the sorting accuracy rate for a future time period; the waste transfer station model outputs predictions of equipment failure probability and transfer efficiency; and the waste disposal site model provides predictions of improved treatment effectiveness.

[0060] The loss value is calculated using mean squared error based on the difference between the predicted results and the true labels.

[0061] The gradients corresponding to the local parameters are calculated based on the loss value using the backpropagation algorithm. Using the calculated loss value, the gradients for each local model parameter are determined by calculating the gradients from the output layer to the input layer using the backpropagation algorithm.

[0062] Based on the gradient, stochastic gradient descent is used to update the parameters of the local model, completing one local model training iteration. This process is repeated multiple times until the preset training stopping condition is met.

[0063] S24: The federated learning server uses a secure aggregation algorithm to aggregate and calculate encrypted gradient information, obtaining updated values ​​for the parameters of the city status assessment model and updating the model. Through this algorithm, encrypted gradient information from various local nodes (including nodes related to different business operations such as waste sorting and disposal points, waste transfer stations, and waste disposal facilities) is summarized and aggregated. Taking into account the different aspects reflected in the data from each node, updated values ​​for the overall parameters of the city status assessment model are derived. This ensures that the model is optimized and updated without compromising the privacy of the original data of each node, enabling a more comprehensive and accurate assessment of the city's operational status in waste sorting and disposal.

[0064] S25: Input the new equipment perception data and historical operation and maintenance records into the trained urban status assessment model and deploy it online. Output the urban operation status assessment results through model inference, including the facility health of each management unit (whether the transfer equipment of the garbage transfer station is in good operating condition, whether the facilities of the garbage sorting and disposal points are intact, and whether the key equipment of the garbage harmless treatment workshop is operating normally), the urgency of service needs (excessive garbage accumulation at the garbage transfer station urgently requires the allocation of vehicles for transfer, the continuous decline in the accuracy of residents' sorting at the garbage sorting and disposal points requires urgent strengthening of publicity and guidance, and the abnormality of the garbage harmless treatment process requires immediate adjustment), and resource matching rationality indicators (whether the allocation of personnel, vehicles and other resources at the garbage transfer station can meet the current transfer volume demand, whether the number of sorting guidance personnel, disposal points and residents in the operation of the garbage sorting project are matched, and whether the allocation of material and equipment resources in each link of the garbage harmless treatment is reasonable).

[0065] S3: Based on the assessment results, combined with service standards and resource allocation, urban management needs are broken down into multiple service tasks, and a service response matrix is ​​constructed by linking device perception data and operation and maintenance resources.

[0066] S31: Extract core management needs based on the city status assessment results, clarifying the scope of the needs (whether it's a specific residential area, commercial area, or the entire street's waste sorting and disposal issues), the degree of impact (e.g., the impact of a waste transfer station malfunction causing waste accumulation on the surrounding environment, sanitation, and traffic; the impact of incorrect waste sorting at collection points on the overall progress of waste sorting in the area), and the urgency level (e.g., a complete shutdown of a waste transfer station due to equipment failure is an emergency, while occasional minor mis-sorting at collection points is a normal situation). Simultaneously, compare these needs with pre-set city service standards (e.g., accuracy standards for waste sorting, etc.). The core management needs list and corresponding service requirements are determined by establishing transfer frequency standards for transfer stations and various environmental protection standards for harmless waste treatment. This ensures a clear and definite basis for subsequent task breakdown and resource allocation. For example, for waste sorting collection points, this might include sorting guidance and promotion services; for waste transfer stations, it might include equipment maintenance and transfer scheduling services; and for harmless waste treatment, it might include process optimization and equipment upgrade services. The quality indicators (such as the required improvement in residents' sorting accuracy after guidance, the guaranteed trouble-free operation time after equipment maintenance, and the environmental protection standards to be achieved after harmless treatment process optimization) and completion deadlines are also defined.

[0067] S32: Based on the currently available operation and maintenance resources, including personnel skills and qualifications (such as publicity and guidance personnel with professional knowledge of waste sorting and good communication skills in the operation and management of waste sorting projects, technical workers with equipment maintenance skills and operation qualifications in waste transfer stations, and professional engineers familiar with various harmless treatment processes in waste harmless treatment), equipment types and status (such as the condition of transfer vehicles and compression equipment of different tonnages in waste transfer stations, whether the intelligent sorting facilities and monitoring equipment at waste sorting points are operating normally, and the operating status of incineration equipment and landfill facilities for waste harmless treatment), material reserves (such as the quantity of garbage bags and disinfectants in waste transfer stations, the quantity of publicity materials for waste sorting projects, and the reserves of chemical agents required for waste harmless treatment), and spatial resource distribution (such as the garbage storage space and vehicle parking space in waste transfer stations, the reasonable layout space of waste sorting points, and the site space of each workshop in waste harmless treatment), and in conjunction with the service types involved in the core management requirements list, establish a correspondence between resource attributes and service capabilities.

[0068] S33: Based on the determined service type, quality indicators, and completion deadlines, the core requirements are broken down into several interrelated sub-tasks according to the service process and professional division of labor. Each sub-task clearly defines the specific operational content, the type of executing entity, and the required resource conditions, and the required resource conditions must correspond to the resource attributes. In a specific embodiment, the equipment maintenance service of a waste transfer station can be broken down into equipment fault diagnosis sub-tasks, parts replacement sub-tasks, and equipment debugging sub-tasks. These sub-tasks are interconnected, and the completion status of the previous sub-task affects the progress of subsequent tasks. Furthermore, the resources required for each sub-task are consistent with the actual resource attributes, ensuring the smooth operation of the entire maintenance service.

[0069] S34: Establish an initial matrix using subtasks and resources, adjust the initial matrix using data adaptation and priority, and obtain the service response matrix.

[0070] S341: Construct an initial matrix with the decomposed subtasks as rows and maintenance resources as columns. Based on the task location reflected in the equipment perception data (such as the specific site location of the waste transfer station equipment maintenance task, the specific collection point location corresponding to the waste sorting and collection point guidance task, etc.), the status of the facilities to be processed (such as the specific fault condition of the waste transfer station equipment to be repaired, whether the waste sorting and collection point facilities are damaged, etc.), combined with the geographical location of resources (such as the current location of maintenance technicians, the location of the spare parts storage warehouse, the distribution location of publicity and guidance personnel, etc.), skill matching degree (whether the personnel skills meet the task requirements, such as whether the maintenance personnel are familiar with the technology of the equipment to be repaired, whether the publicity personnel have mastered the relevant waste sorting knowledge, etc.), and real-time load status (such as whether the maintenance personnel currently have other maintenance tasks, whether the vehicles have been dispatched and used), mark the resource and task adaptability and response priority in the matrix cells. The adaptability assessment should refer to the resource conditions required by the subtask.

[0071] S342: Based on the service standard timeliness requirements and the principle of optimal resource allocation, the initial matrix is ​​adjusted to remove resource-task combinations with a suitability below the threshold, strengthen the association between high-priority tasks and high-quality resources, and dynamically correct resource status parameters through real-time updates of device perception data to ensure that the resource information in the matrix corresponds to the resource attributes.

[0072] The output is an adjusted service response matrix. Each valid cell in the matrix contains a task identifier, a resource identifier, an estimated response time, and a service quality commitment, enabling precise correlation between device perception data and operation and maintenance resources, and providing a direct basis for task allocation.

[0073] S4: Calculate the priority coefficient of each service task in the service response matrix and dynamically schedule resources to generate the optimal service solution.

[0074] S41: Calculate the priority coefficient of each service task in the service response matrix, which is achieved through the following sub-steps: S411: Based on the attribute information of each service task in the service response matrix, extract the core parameters affecting priority, including the urgency level, scope of impact, importance of associated facilities, and current unprocessed duration of the task. The urgency level and scope of impact are derived from the core management requirements list, and the facility importance is associated with the facility type information in the equipment perception data.

[0075] S412: Assign weight values ​​to each parameter, with the urgency level having the highest weight, followed by facility importance, then the scope of impact, and finally the unprocessed duration. Calculate the priority coefficient for each service task by weighted summation.

[0076] In a specific embodiment, the formula for calculating the priority coefficient is as follows: Priority coefficient = ω1 × Emergency level score + ω2 × Facility importance score + ω3 × Impact range score + ω4 × Current unprocessed time score.

[0077] For the emergency repair task of the garbage transfer station, the urgency level is "high", with a corresponding score of 0.8. The facility importance is also "high", with a corresponding score of 0.9. The impact range is "large", with a corresponding score of 0.7. The current unprocessed time is 1 day, with a corresponding score of 0.3. Different score ranges can be divided according to the time, such as 0-1 days for 0.3, 1-3 days for 0.5, etc. Then the priority coefficient of this task is 0.76.

[0078] S413: Sort all service tasks according to priority coefficients; the higher the coefficient, the higher the task priority.

[0079] S42: Dynamically schedule resources based on priority coefficients and generate the optimal service plan. The specific steps are as follows: Based on the resource and task compatibility and real-time resource status in the service response matrix, the system matches the best resources to tasks according to their priority, prioritizing the allocation of resources with high compatibility and low current load to high-priority tasks, while also considering the geographical distribution of resources to shorten response time.

[0080] In the resource allocation process, if resource conflicts occur, specifically when the same group of experienced maintenance experts are simultaneously needed for multiple high-priority equipment maintenance tasks, or when a garbage transfer vehicle in good condition and with high adaptability is simultaneously called upon for transfer tasks in different areas, the priority coefficients of the conflicting tasks are compared. The task with the higher priority coefficient is retained in the association with the resource, and the remaining tasks are reassigned to the next best resource. For example, if two garbage transfer stations require the same maintenance expert for equipment maintenance tasks, and the priority coefficient for the equipment repair task at transfer station A is 0.8, while the priority coefficient for the equipment repair task at transfer station B is 0.6, then the maintenance expert will be prioritized for the task at transfer station A. For the task at transfer station B, other maintenance personnel with slightly less experience but still capable of handling the task will be found, or the expert will be reassigned after completing the task at transfer station A. At the same time, other available equipment, spare parts, and other resources will be allocated to ensure the basic operation of transfer station B during the waiting period.

[0081] Finally, based on the resource allocation results, an optimal service plan is generated, which includes the task execution order, resource scheduling path, and estimated completion time. The plan must meet the quality indicators and time limits in the service standards. At the same time, the resource status and task priority coefficients are dynamically adjusted through real-time updates of device perception data to continuously optimize the service plan.

[0082] In summary, this embodiment provides an Internet of Things-based urban management service method. By fusing multi-source sensing data through a federated learning algorithm to construct an urban status assessment model, it solves the shortcomings of low data utilization in urban management, which makes it difficult to achieve refined and intelligent urban management.

[0083] By leveraging federated learning algorithms to integrate multi-source sensing data to construct an urban status assessment model, and fully integrating scattered information resources, this model can not only comprehensively consider the multi-dimensional situation of each management unit, but also output key indicators such as facility health and the urgency of service needs. This allows managers to clearly understand the city's operational status from a macro perspective, so as to make practical and holistic planning and control measures.

[0084] Based on the city status assessment results, management needs are broken down into multiple service tasks, and a service response matrix is ​​constructed by linking device perception data and operation and maintenance resources. This enables the rational allocation of resources according to the actual situation, prioritizing important and urgent tasks, avoiding resource waste, improving resource utilization efficiency, and maximizing the value of limited human, material, and equipment resources to ensure the efficient and orderly operation of city services.

[0085] Based on the same general inventive concept, this invention also protects an Internet of Things (IoT)-based urban concierge service method system. The following describes the IoT-based urban concierge service method system provided by this invention. The IoT-based urban concierge service method system described below can be referred to in correspondence with the IoT-based urban concierge service method described above.

[0086] Figure 2 This is a schematic diagram of the structure of an Internet of Things-based urban concierge service system provided in an embodiment of the present invention.

[0087] like Figure 2 As shown, an Internet of Things (IoT) based urban concierge service system includes: The sensing data acquisition module is used to divide management units according to urban area characteristics, deploy IoT sensing devices to collect urban element data, define facility sensing templates based on management unit functions, and classify sensing devices according to monitoring type to obtain device sensing data.

[0088] The city operation status assessment module uses a federated learning algorithm to integrate multi-source sensing data to build a city status assessment model. It takes device sensing data and historical operation and maintenance records as inputs and generates city operation status assessment results.

[0089] The matrix construction module breaks down urban management needs into multiple service tasks based on urban operation status assessment results, service standards, and resource allocation, and constructs a service response matrix by associating device perception data with operation and maintenance resources.

[0090] The solution generation module calculates the priority coefficient of each service task in the service response matrix and dynamically schedules resources to generate the optimal service solution.

[0091] This invention provides an Internet of Things (IoT)-based urban management service method that collects various types of urban management data in real time. This solves the problem of low data utilization and difficulty in achieving refined and intelligent urban management across all aspects of urban management.

[0092] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0093] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for providing urban concierge services based on the Internet of Things, characterized in that, include: S1: Divide the city into management units based on the characteristics of the urban area, deploy IoT sensing devices to collect urban element data, define facility sensing templates based on the functions of the management units, and classify the sensing devices according to the monitoring type to obtain device sensing data; S2: Based on federated learning algorithms, a city status assessment model is constructed by fusing multi-source sensing data. Input device sensing data and historical operation and maintenance records to generate city operation status assessment results. S3: Based on the city operation status assessment results, service standards and resource allocation, the city management needs are broken down into multiple service tasks, and a service response matrix is ​​constructed by associating the device perception data and operation and maintenance resources. S4: Calculate the priority coefficient of each service task in the service response matrix, and dynamically schedule resources to generate the optimal service solution.

2. The method for providing urban concierge services based on the Internet of Things according to claim 1, characterized in that, In step S1, the specific steps for acquiring device sensing data are as follows: S11: Divide management units according to urban functional areas and geographical boundaries, label the core facility type in each unit, and determine perception indicators based on the characteristics of facilities within the unit; S12: Based on the core facility type and perception indicators, feature extraction is performed on the management unit to obtain core service requirements. Perception dimensions are set for each facility, and application scenarios are defined according to the core service requirements to obtain facility perception templates. S13: Collect the raw data of the sensing devices in the facility sensing template, and use image recognition technology to convert the unstructured data in the raw data into structured raw data corresponding to the sensing indicators; S14: The structured raw data is classified according to the detection type to obtain device sensing data.

3. The method for providing urban concierge services based on the Internet of Things according to claim 2, characterized in that, In step S14, the specific steps for classifying the structured raw data are as follows: Extract the perception dimensions corresponding to the structured indicators from the facility perception template; The weights of the perception dimensions are calculated based on the importance of the facilities in the management unit, and the structured indicators are compared with the facility perception template to obtain the equipment data quality score. The overall effectiveness score of the device data is calculated based on the weights of the perception dimensions and the device data quality score. The structured raw data is then divided according to the score to obtain the device perception data.

4. The method for providing urban concierge services based on the Internet of Things according to claim 1, characterized in that, In step S2, the specific steps to obtain the urban operation status assessment results are as follows: S21: Clean the equipment sensing data and historical operation and maintenance records, remove outliers, supplement missing values ​​using time series interpolation, extract time and spatial features, and obtain preprocessed sensing data; S22: Construct a federated learning framework based on the preprocessed sensory data, and use a deep neural network as the local model for federated learning; S23: Local nodes use their own device perception data and historical operation and maintenance records to train local models, calculate the update gradient of local model parameters, and upload the encrypted gradient information to the federated learning server to build a city status assessment model. S24: The federated learning server uses a secure aggregation algorithm to aggregate and calculate encrypted gradient information, obtain updated values ​​of the parameters of the city status assessment model, and update the city status assessment model. S25: Input the new equipment perception data and historical operation and maintenance records into the trained urban status assessment model, and output the urban operation status assessment results through model inference.

5. A method for providing urban concierge services based on the Internet of Things according to claim 4, characterized in that, In step S23, the specific steps for training the local model at the local node are as follows: Initialize the parameters of the deep neural network model; The preprocessed sensing data is input into the local model, and forward propagation calculation is performed according to the gradient descent method to obtain the prediction result; Based on the difference between the predicted results and the true labels, the loss value is calculated using the mean squared error. The gradient corresponding to the local parameters is calculated based on the loss value using the backpropagation algorithm; Based on the gradient, stochastic gradient descent is used to update the parameters of the local model, completing one local model training iteration. This process is repeated multiple times until a preset training stopping condition is met.

6. The method for providing urban concierge services based on the Internet of Things according to claim 1, characterized in that, In step S3, the specific steps for constructing the service response matrix are as follows: S31: Extract core management needs based on the city status assessment results to obtain a list of core management needs; S32: Based on the currently available operation and maintenance resources and the service types in the core management requirements list, establish the correspondence between resource attributes and service capabilities; S33: Based on the service type, the core requirements are broken down into several interrelated sub-tasks according to the service process and professional division of labor; S34: Establish an initial matrix using subtasks and resources, adjust the initial matrix using data adaptation and priority, and obtain the service response matrix.

7. A method for providing urban concierge services based on the Internet of Things according to claim 6, characterized in that, In step S34, the specific steps for adjusting the service response matrix are as follows: S341: Construct an initial matrix with subtasks as rows and operation and maintenance resources as columns, and evaluate the adaptability based on adaptability and priority to obtain the initial matrix; S342: Update and correct resource parameters according to service standards and resource configuration, and adjust the initial matrix.

8. A method for providing urban concierge services based on the Internet of Things according to claim 1, characterized in that, In step S4, the specific steps for calculating the priority coefficient of each service task in the service response matrix are as follows: Based on the attribute information of each service task in the service response matrix, the core parameters affecting priority are extracted, a weight value is assigned to each parameter, the priority coefficient of each service task is calculated by weighted summation, and all service tasks are sorted according to the priority coefficient.

9. A method for providing urban concierge services based on the Internet of Things according to claim 1, characterized in that, In step S4, the specific steps for dynamically scheduling resources based on priority coefficients and generating the optimal service plan are as follows: Based on the resource and task compatibility and real-time resource status in the service response matrix, the resources are sorted according to the priority coefficient to match the optimal resources for the tasks. Resources with high compatibility and low current load are allocated to high-priority tasks first. At the same time, the geographical distribution of resources is considered to shorten the response time. The optimal service plan is generated based on the resource allocation results.

10. An Internet of Things (IoT)-based urban concierge service system, applied to an IoT-based urban concierge service method as described in any one of claims 1 to 9, characterized in that, The butler service system includes: The sensing data acquisition module is used to divide management units according to urban area characteristics, deploy IoT sensing devices to collect urban element data, define facility sensing templates based on management unit functions, and classify sensing devices according to monitoring type to obtain device sensing data. The city operation status assessment module uses a federated learning algorithm to integrate multi-source sensing data to build a city status assessment model. It takes device sensing data and historical operation and maintenance records as inputs and generates city operation status assessment results. The matrix construction module breaks down urban management needs into multiple service tasks based on the urban operation status assessment results, service standards, and resource allocation, and constructs a service response matrix by associating device perception data with operation and maintenance resources. The solution generation module calculates the priority coefficient of each service task in the service response matrix and dynamically schedules resources to generate the optimal service solution.

Images