A smart environmental sanitation management method and device, computer equipment and readable medium

Abstract

This invention discloses a smart sanitation management method, device, computer equipment, and readable medium, relating to the field of sanitation management technology. The method includes: constructing a relational database between influencing factor entities and sanitation problem entities; acquiring actual event data for each sub-area within the sanitation area; calling the relational database, combining the current event data to predict risk indicators for sanitation problem entities in each sub-area, and filtering sanitation problem entities with risk indicators exceeding a threshold as pending sanitation problems; determining the risk indicator intensity for each type of pending sanitation problem based on sanitation management experience; and dynamically generating and implementing a cleaning intensity adjustment plan for the corresponding sub-area based on the risk indicator intensity of the pending sanitation problem type. The advantages of this invention are: by constructing a dynamic intelligent correlation model between events and sanitation problems, it reduces urban sanitation operation costs and maintains a consistently high level of environmental sanitation quality.

Description

Technical Field

[0001] This invention relates to the field of sanitation management technology, specifically to a smart sanitation management method, device, computer equipment, and readable medium. Background Technology

[0002] Traditional urban sanitation management relies primarily on fixed shifts, manual patrols, and experience-based judgment to maintain sanitation. This approach suffers from significant delays and resource misallocation. Faced with a highly dynamic urban environment, such as the non-constant distribution of pedestrian traffic in time and space, sudden garbage generation, and the impact of extreme weather events, static management models struggle to respond quickly and allocate resources accurately. Consequently, sanitation problems are often only discovered and addressed after they have worsened or spread, leading to high cleaning costs, fluctuating environmental quality, and difficulty in guaranteeing public satisfaction.

[0003] Existing technologies attempt to introduce IoT sensors to monitor waste volume or utilize basic data to analyze the impact of single factors on sanitation. However, these methods typically have limitations: first, they fail to systematically identify and quantify the multidimensional correlations between various dynamic events and complex sanitation problems; second, they lack a comprehensive assessment of the risks of mutual induction and transmission among sanitation problems; and third, resource allocation strategies often neglect real-time risk intensity and global optimization constraints, easily leading to over- or under-response in certain areas. Therefore, there is an urgent need for a smart sanitation management solution that can integrate multi-source data, gain insights into the deep correlations between events and problems, scientifically quantify risks, and dynamically optimize cleaning efforts to achieve precise, efficient, and preventative resource allocation. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a smart sanitation management method, device, computer equipment, and readable medium. This technical solution resolves at least one of the technical problems existing in the prior art mentioned in the background section.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A smart sanitation management method includes:

[0007] Several dynamic event types affecting regional sanitation conditions are identified as influencing factor entities, and the types of sanitation problems occurring in the sanitation area are identified as sanitation problem entities. A database of correlations between influencing factor entities and sanitation problem entities is constructed based on historical sanitation data.

[0008] Acquire actual event data for each sub-area within the sanitation area, including at least one of pedestrian traffic, garbage generation, and abnormal weather events;

[0009] Call the relational database, combine it with the current event data to predict the risk indicators of entities with health problems in each sub-region, and filter out health problem entities whose risk indicators exceed the threshold and record them as health problems to be handled;

[0010] Based on sanitation management experience, determine the risk indicator intensity for each type of sanitation problem to be addressed;

[0011] Based on the risk index intensity of the type of hygiene problem to be addressed, a cleaning intensity adjustment plan for the corresponding sub-area is dynamically generated and implemented.

[0012] Preferably, the construction of the relational database between influencing factor entities and sanitation problem entities based on historical sanitation data specifically includes:

[0013] Extract event-problem association rules from historical sanitation data and construct a triplet knowledge base with influencing factor entities as head nodes, sanitation problem entities as tail nodes, and association relationships as edges;

[0014] Based on the triplet knowledge base, a corresponding risk prediction model is established for each health problem entity. The risk prediction model adopts a logistic regression model or a neural network model.

[0015] Preferably, when the risk prediction model is a neural network model, the process of establishing the risk prediction model is as follows:

[0016] The historical event data was labeled as the training set according to the occurrence status of the health issues;

[0017] The neural network is trained with event parameters as the input layer and the probability of problem occurrence as the output layer.

[0018] The accuracy of the model is verified using a test set, and the model with the smallest loss function value is selected as the final risk prediction model.

[0019] Preferably, the step of calling the relational database and combining it with current event data to predict the risk indicators of entities with health problems in each sub-region specifically involves:

[0020] For each influencing entity, sample data at multiple time scales are collected;

[0021] Predict future event parameter values ​​at various scales using linear regression;

[0022] Weighted fusion of multi-scale prediction values ​​yields comprehensive prediction parameters;

[0023] The comprehensive prediction parameters are input into the risk prediction model, and the predicted risk index is output.

[0024] Preferably, the determination of the risk indicator intensity for each type of sanitation problem to be addressed based on sanitation management experience specifically includes:

[0025] Based on sanitation management experience, a sanitation problem association graph is constructed, in which sanitation problem types are used as graph nodes and the inducing relationships between sanitation problems are used as directed edges;

[0026] Based on historical sanitation management experience and the required processing costs for each type of sanitation problem, the intensity of the risk index for each type of sanitation problem is calculated.

[0027] Based on the risk index intensity of the health problem type that will be induced by the failure to address the health problem type in a timely manner, a PAGERANK-like algorithm is used to calculate the risk index intensity of each health problem type to be addressed.

[0028] Preferably, the step of dynamically generating and implementing a cleaning intensity adjustment plan for the corresponding sub-area based on the risk indicator intensity of the type of sanitation problem to be addressed specifically includes:

[0029] Based on sanitation experience, determine the optimal sanitation resource requirements for each sanitation problem to be addressed.

[0030] Given the constraints that the sanitation resources allocated to each sanitation problem to be addressed are less than or equal to the optimal sanitation resource requirements of the sanitation problem to be addressed, and that the sum of the sanitation resources allocated to all sanitation problems to be addressed is less than the total amount of sanitation resources, several alternative sanitation allocation schemes are generated.

[0031] Based on the similarity between the sanitation resources allocated to each sanitation problem in the candidate sanitation allocation scheme and the optimal sanitation resource demand of the sanitation problem, the sanitation resources are used as the treatment intensity of the sanitation problem. The treatment intensity of the sanitation problem is multiplied by the risk index intensity of the sanitation problem type to obtain the treatment index of the sanitation problem.

[0032] The total problem-solving index for the candidate sanitation allocation scheme is obtained by summing up the processing indicators of all pending sanitation problems.

[0033] The candidate sanitation allocation scheme with the minimum total problem handling index is selected as the optimal sanitation allocation scheme.

[0034] Based on the optimal sanitation allocation plan and the current sanitation resource allocation, using the quantity and time of sanitation resource mobilization as evaluation factors, the TOPSIS method is used to screen out the optimal cleaning intensity adjustment plan that transforms the current sanitation resource allocation into the optimal sanitation allocation plan.

[0035] Furthermore, a smart sanitation management system is proposed to implement the smart sanitation management method described above, including:

[0036] The data acquisition module is used to acquire actual event data of each sub-area within the sanitation area. The event data includes at least one of the following: pedestrian flow, garbage generation, and abnormal weather events.

[0037] The knowledge base construction module is used to identify several dynamic event types that affect the regional sanitation status as influencing factor entities, determine the types of sanitation problems that occur in the sanitation area as sanitation problem entities, and build a relational database between influencing factor entities and sanitation problem entities based on historical sanitation data.

[0038] The risk prediction module connects the knowledge base construction module and the data acquisition module. It is used to call the relational database and combine it with the current event data to predict the risk indicators of health problem entities in each sub-region, and to screen health problem entities whose risk indicators exceed the threshold as health problems to be handled.

[0039] The risk intensity calculation module, connected to the risk prediction module, is used to determine the risk index intensity of each type of sanitation problem to be addressed based on sanitation management experience.

[0040] The cleaning scheduling engine, connected to the risk intensity calculation module, is used to dynamically generate and execute cleaning intensity adjustment plans for corresponding sub-areas based on the risk index intensity of the type of hygiene problem to be addressed.

[0041] Optionally, the knowledge base construction module includes:

[0042] The rule extraction unit is used to extract event-problem association rules from historical sanitation data;

[0043] The triplet storage unit is used to construct a triplet knowledge base with influencing factor entities as head nodes, health problem entities as tail nodes, and relationships as edges;

[0044] The model training unit is used to build a corresponding Logistic regression model or neural network model for each health problem entity as a risk prediction model.

[0045] The clean scheduling engine includes:

[0046] The resource demand mapping unit is used to determine the optimal sanitation resource demand for each sanitation problem to be addressed based on sanitation experience.

[0047] The scheme generation unit is used to generate several candidate sanitation allocation schemes with the constraints that allocated resources ≤ optimal resource demand and total resources ≤ total sanitation volume.

[0048] The processing index calculation unit is used to calculate the cumulative value of (allocated resources / optimal resource requirements) × risk index intensity in each candidate solution as the overall problem processing index;

[0049] The scheme selection unit is used to select the candidate scheme with the smallest total problem handling index as the best sanitation allocation scheme.

[0050] The TOPSIS scheduling unit is used to generate an optimal cleaning intensity adjustment plan from the current allocation state to the best allocation plan by using the quantity and time of sanitation resource mobilization as evaluation factors and employing the TOPSIS method.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0052] This invention significantly improves the foresight and accuracy of sanitation risk prediction by constructing an intelligent correlation model between dynamic events and sanitation issues. By systematically integrating multi-source real-time data and historical management experience, it not only accurately identifies high-probability sanitation problem types and their potential cascading risks, but also dynamically generates optimal cleaning scheduling strategies based on global constraints. Ultimately, this achieves a shift from passive response to proactive prevention, minimizing the spread and deterioration of sanitation problems while maximizing the allocation of sanitation resources, reducing overall urban sanitation operating costs, and maintaining a consistently high level of environmental sanitation quality. Attached Figure Description

[0053] Figure 1 Here is a flowchart of the intelligent sanitation management method proposed in Example 1;

[0054] Figure 2 Here is a flowchart of the method for constructing a relational database between influencing factor entities and health problem entities, as proposed in Example 2.

[0055] Figure 3 This is a flowchart of the method for determining the intensity of risk indicators for each type of health problem to be addressed, as proposed in Example 3.

[0056] Figure 4 This is a flowchart of the method for dynamically generating cleaning intensity adjustment schemes for corresponding sub-regions as proposed in Example 3;

[0057] Figure 5 This is an architecture diagram of the electronic devices in this solution;

[0058] Figure 6 This is a schematic diagram of the computer-readable storage medium structure in this scheme.

[0059] The numbers on the map are:

[0060] 500 - Electronic device; 501 - Bus; 502 - CPU; 503 - ROM; 504 - RAM; 505 - Communication port; 506 - Input / output component; 507 - Hard disk; 508 - User interface; 600 - Computer-readable storage medium. Detailed Implementation

[0061] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0062] Reference Figure 1 As shown, a smart sanitation management method includes:

[0063] This method identifies several dynamic event types affecting regional sanitation as influencing factor entities and determines the types of sanitation problems occurring in sanitation areas as sanitation problem entities. Based on historical sanitation data, it constructs a relational database between influencing factor entities and sanitation problem entities. By breaking through the traditional experience-dependent model, it transforms fragmented sanitation knowledge into a quantifiable entity relational network through structured historical data. This eliminates biases in subjective human judgment, systematically establishes causal relationship chains between events and sanitation problems, provides a highly reliable data foundation for subsequent intelligent prediction, and significantly improves the comprehensiveness and interpretability of sanitation risk identification.

[0064] The system acquires real-time event data from various sub-areas within the sanitation area. This data includes at least one of the following: pedestrian traffic, waste generation, and abnormal weather events. Leveraging the Internet of Things (IoT) and multi-source sensing technology, it achieves refined regional-level data collection, overcoming the blind spots of traditional fixed-point monitoring. It captures key variables such as sudden changes in pedestrian traffic, dynamic waste accumulation, and the impact of severe weather in real time, providing high spatiotemporal resolution input for risk early warning.

[0065] By accessing a relational database and combining current event data, the system predicts risk indicators for entities with health problems in each sub-region. Entities with risk indicators exceeding a threshold are identified as health problems requiring immediate attention. By integrating real-time data with historical association rules, the system achieves probabilistic and accurate prediction of health problems. A machine learning model automatically filters high-risk problem types, replacing the inefficient and blind testing method of manual inspections. This allows health management resources to prioritize core issues that are about to erupt, preventing the spread of risks.

[0066] Based on sanitation management experience, the risk indicator intensity of each type of sanitation problem to be addressed is determined, the chain reaction effect of sanitation problems is quantified, and the impact of a single event is expanded into a multi-level risk network assessment. By calculating the global impact weight of problem types, the misjudgment of the priority of local problems in traditional methods is avoided, ensuring that sanitation problems with high infectivity and high treatment costs receive priority.

[0067] Based on the risk index intensity of the types of sanitation problems to be addressed, a corresponding sub-area cleaning intensity adjustment plan is dynamically generated and executed, upgrading the static cleaning plan to a dynamic resource optimization model. Under the constraint of limited sanitation resources, the optimal scheduling strategy is generated. By matching the problem risk intensity and the handling intensity in real time, the contradiction between resource waste and insufficient response is avoided, thereby improving overall cleaning efficiency and reducing operating costs.

[0068] Reference Figure 2 As shown, the database of relationships between influencing factor entities and sanitation problem entities, constructed based on historical sanitation data, specifically includes:

[0069] Extract event-problem association rules from historical sanitation data and construct a triplet knowledge base with influencing factor entities as head nodes, sanitation problem entities as tail nodes, and association relationships as edges;

[0070] Based on the triplet knowledge base, a corresponding risk prediction model is established for each health problem entity. The risk prediction model adopts either a logistic regression model or a neural network model.

[0071] Specifically, in some embodiments, when the risk prediction model uses a Logistic regression model, the expression for the risk prediction model is:

[0072]

[0073] In the formula, This represents the probability value of entities with hygiene problems appearing in the sub-region. Let be the comprehensive prediction parameter for the i-th future event in the sub-region, and be the total number of influencing entity entities. All of these are model parameters, obtained using the least squares method;

[0074] Specifically, the Logistic Regression model is a generalized linear regression analysis model, commonly used in data mining, automated disease diagnosis, economic forecasting, and other fields. The Logistic Regression model estimates the probability of an event occurring based on a given dataset of independent variables. Since the result is a probability, the dependent variable ranges between 0 and 1.

[0075] In other embodiments, when the risk prediction model is a neural network model, the process of establishing the risk prediction model is as follows:

[0076] The historical event data was labeled as the training set according to the occurrence status of the health issues;

[0077] The neural network is trained with event parameters as the input layer and the probability of problem occurrence as the output layer.

[0078] The accuracy of the model is verified using a test set, and the model with the smallest loss function value is selected as the final risk prediction model.

[0079] This embodiment achieves a visual causal mapping between events and issues by constructing a triplet knowledge base, providing interpretable rule support for risk modeling.

[0080] When the Logistic regression model is used in some implementations, its closed mathematical expression can clearly quantify the influence weight of different event parameters on the probability of health problems, and can be quickly deployed in lightweight scenarios with limited computing power.

[0081] In other embodiments, neural network models are used to automatically capture implicit association rules of complex event combinations through deep nonlinear fitting capabilities, overcoming the limitations of traditional statistical models in representing high-dimensional dynamic data. The dual-mode architecture design ensures both the traceability of prediction results in simple scenarios and the adaptive improvement of prediction accuracy in extremely complex environments, thereby comprehensively enhancing the robustness and scenario generalization ability of risk warning.

[0082] In some preferred embodiments, a combination of Logistic regression and neural network models can be used to predict the probability of occurrence of health problem entities, thereby effectively improving the prediction accuracy of health problem entities.

[0083] The specific steps for predicting risk indicators for entities with health problems in each sub-region by calling upon the relational database and combining it with current event data are as follows:

[0084] For each influencing entity, sample data at multiple time scales are collected;

[0085] Predict future event parameter values ​​at various scales using linear regression;

[0086] Weighted fusion of multi-scale prediction values ​​yields comprehensive prediction parameters;

[0087] The comprehensive prediction parameters are input into the risk prediction model, and the predicted risk index is output.

[0088] Specifically, regional event parameters typically exhibit varying cyclical fluctuations. For instance, regional pedestrian traffic is related not only to the time of day but also to whether it's a weekday, involving two different scales of cyclical fluctuations: daily and weekly. Similarly, regional leaf fall is related not only to the real-time tree population but also to the season. Therefore, analyzing regional leaf fall requires simultaneously analyzing daily leaf fall to assess the tree population and also analyzing leaf fall during the same season each year. This again involves two different scales of cyclical fluctuations: daily and annual. Furthermore, analyzing sudden events... When analyzing event parameter values, such as changes in visitor traffic caused by a concert, it is necessary to refer to historical event reference values ​​for similar emergencies. Therefore, this solution analyzes future event parameter values ​​at multiple different scales and combines them with the main influencing scale to weight and analyze the comprehensive prediction parameters for each time period. For example, when analyzing visitor traffic, during non-long holiday periods, the daily visitor traffic fluctuation is mainly used as the main reference scale, and the weight of this scale is increased. During long holiday periods, the visitor traffic at the same time in previous years is mainly used as the main reference scale, and the annual daily visitor traffic fluctuation is used as the main reference scale, and the weight of the annual scale is increased.

[0089] This multi-timescale coupled prediction mechanism addresses the challenge of cyclical confounding in dynamic events. For parameters exhibiting multiple fluctuations, such as pedestrian traffic and leaf fall, it simultaneously analyzes evolution patterns at daily, weekly, yearly, and event-driven scales, avoiding misjudgments by single-scale models during long holidays, seasonal transitions, or sudden events. Its dynamic weighted fusion strategy adaptively strengthens the weight of the dominant scale; for example, it emphasizes short-cycle fluctuation patterns during normal times while anchoring to long-cycle historical benchmarks during long holidays, ensuring that the comprehensive prediction parameters possess both short-term sensitivity and long-term robustness. Ultimately, it significantly reduces prediction blind spots in scenarios such as holiday effects and abrupt climate changes, providing high-fidelity input for subsequent risk models and reducing false positive and false negative rates for health issues.

[0090] Example 3:

[0091] Reference Figure 3 As shown, based on sanitation management experience, the specific risk indicator intensity for each type of sanitation problem to be addressed includes:

[0092] Based on sanitation management experience, a sanitation problem association graph is constructed, in which sanitation problem types are used as graph nodes and the inducing relationships between sanitation problems are used as directed edges;

[0093] Based on historical sanitation management experience and the required processing costs for each type of sanitation problem, the intensity of the risk index for each type of sanitation problem is calculated.

[0094] Based on the risk index intensity of the health problem type that will be induced by the failure to address the health problem type in a timely manner, a PAGERANK-like algorithm is used to calculate the risk index intensity of each health problem type to be addressed.

[0095] The calculation process for the risk indicator intensity of the specific types of health problems to be addressed is as follows:

[0096] First, the risk index intensity of all health problem types is normalized to obtain the initial PR value of all health problem types;

[0097] The initial PR values ​​for all health problem types are substituted into the PAGERANK formula to obtain the analytical PR value for each health problem type. After normalizing the analytical PR values, the updated PR values ​​for each health problem type are obtained. Then, the updated PR values ​​for all health problem types are substituted into the PAGERANK formula for iterative iteration until the updated PR values ​​for all health problem types converge.

[0098] Extract the convergence value of the updated PR value of the health problem type to be addressed, and use it as the risk indicator strength of the health problem type to be addressed.

[0099] Specifically, the PageRank formula is:

[0100]

[0101] In the formula, Let be the PR value for the i-th health problem type. Let be the set of health problem types that will be induced if the i-th health problem type is not dealt with in a timely manner. for The elements within it, for Update PR value, The total number of types of health problems. To cause The total number of types of health problems.

[0102] By constructing a health problem association map and integrating an improved PAGERANK algorithm, this approach overcomes the limitations of traditional risk assessment that isolates single health problems. By quantitatively analyzing the intensity of multi-level chain reactions induced by untreated health problems, it deeply identifies key risk hubs with high transmission potential. Its dynamic iterative calculation process accurately captures the global propagation effect of risks in complex networks, enabling resources to prioritize blocking health problem nodes that may lead to systemic deterioration. This significantly improves the foresight and resilience of urban sanitation management, reducing the incidence of chain public health events while optimizing the efficiency of emergency resource allocation.

[0103] Reference Figure 4As shown, based on the risk indicator intensity of the type of sanitation problem to be addressed, a cleaning intensity adjustment plan for the corresponding sub-area is dynamically generated and implemented, specifically including:

[0104] Based on sanitation experience, determine the optimal sanitation resource requirements for each sanitation problem to be addressed.

[0105] Given the constraints that the sanitation resources allocated to each sanitation problem to be addressed are less than or equal to the optimal sanitation resource requirements of the sanitation problem to be addressed, and that the sum of the sanitation resources allocated to all sanitation problems to be addressed is less than the total amount of sanitation resources, several alternative sanitation allocation schemes are generated.

[0106] Based on the similarity between the sanitation resources allocated to each sanitation problem in the candidate sanitation allocation scheme and the optimal sanitation resource demand of the sanitation problem, the sanitation resources are used as the treatment intensity of the sanitation problem. The treatment intensity of the sanitation problem is multiplied by the risk index intensity of the sanitation problem type to obtain the treatment index of the sanitation problem.

[0107] The total problem-solving index for the candidate sanitation allocation scheme is obtained by summing up the processing indicators of all pending sanitation problems.

[0108] The candidate sanitation allocation scheme with the minimum total problem handling index is selected as the optimal sanitation allocation scheme.

[0109] Based on the optimal sanitation allocation plan and the current sanitation resource allocation, using the quantity and time of sanitation resource mobilization as evaluation factors, the TOPSIS method is used to screen out the optimal cleaning intensity adjustment plan that transforms the current sanitation resource allocation into the optimal sanitation allocation plan.

[0110] Specifically, the TOPSIS process is as follows:

[0111] The ideal optimal scheduling scheme is determined by the minimum number of sanitation resources mobilized and the minimum sanitation resource mobilization time among all global cleaning intensity adjustment schemes.

[0112] The ideal worst-case scheduling scheme is based on the maximum number of sanitation resources mobilized and the maximum sanitation resource mobilization time among all global cleaning intensity adjustment schemes.

[0113] Calculate the normalized vector distance between the number of sanitation resources mobilized and the sanitation resource mobilization time of each cleaning intensity adjustment scheme and the ideal optimal and ideal worst scheduling schemes.

[0114] The TOPSIS value of the cleaning intensity adjustment scheme is obtained by comparing the number and time of sanitation resource mobilization in the cleaning intensity adjustment scheme with the normalized vector distance of the worst-case scheduling scheme with the sum of the normalized vector distances of the number and time of sanitation resource mobilization in the cleaning intensity adjustment scheme with the best-case scheduling scheme and the worst-case scheduling scheme. The cleaning intensity adjustment scheme with the maximum TOPSIS value is selected as the optimal cleaning intensity adjustment scheme.

[0115] A dynamic mapping mechanism between risk intensity and resource demand is established. A global resource allocation scheme is accurately generated through a Pareto optimization model under constraints. The TOPSIS algorithm is used to quickly lock the execution strategy with the highest resource reorganization efficiency among multiple scheduling paths. The decision criterion is to infinitely approximate the ideal solution with the least amount of mobilization and the shortest time. This simultaneously achieves efficient governance of sudden health risks and minimal disturbance to system operation, so that limited sanitation resources can accurately cover key nodes of chain risks and maintain the high-resilience steady-state operation of the urban cleaning service system.

[0116] Example 4:

[0117] To implement the above embodiments one to three, this embodiment proposes a smart sanitation management system, including:

[0118] The data acquisition module is used to acquire actual event data of each sub-area within the sanitation area. The event data includes at least one of the following: pedestrian traffic, amount of garbage generated, and abnormal weather events.

[0119] The knowledge base construction module is used to identify several dynamic event types that affect the regional sanitation status as influencing factor entities, determine the types of sanitation problems that occur in the sanitation area as sanitation problem entities, and build a relational database between influencing factor entities and sanitation problem entities based on historical sanitation data.

[0120] The risk prediction module connects the knowledge base construction module and the data acquisition module. It is used to call the relational database and combine it with the current event data to predict the risk indicators of health problem entities in each sub-region. Health problem entities with risk indicators exceeding the threshold are recorded as health problems to be handled.

[0121] The risk intensity calculation module, connected to the risk prediction module, is used to determine the risk indicator intensity of each type of sanitation problem to be addressed based on sanitation management experience.

[0122] The cleaning scheduling engine, connected to the risk intensity calculation module, is used to dynamically generate and execute cleaning intensity adjustment plans for corresponding sub-areas based on the risk index intensity of the type of hygiene problem to be addressed.

[0123] The knowledge base construction module includes:

[0124] The rule extraction unit is used to extract event-problem association rules from historical sanitation data;

[0125] The triplet storage unit is used to construct a triplet knowledge base with influencing factor entities as head nodes, health problem entities as tail nodes, and relationships as edges;

[0126] The model training unit is used to build a corresponding Logistic regression model or neural network model for each health problem entity as a risk prediction model.

[0127] The clean scheduling engine includes:

[0128] The resource demand mapping unit is used to determine the optimal sanitation resource demand for each sanitation problem to be addressed based on sanitation experience.

[0129] The scheme generation unit is used to generate several candidate sanitation allocation schemes with the constraints that allocated resources ≤ optimal resource demand and total resources ≤ total sanitation volume.

[0130] The processing index calculation unit is used to calculate the cumulative value of (allocated resources / optimal resource requirements) × risk index intensity in each candidate solution as the overall problem processing index;

[0131] The scheme selection unit is used to select the candidate scheme with the smallest total problem handling index as the best sanitation allocation scheme.

[0132] The TOPSIS scheduling unit is used to generate an optimal cleaning intensity adjustment plan from the current allocation state to the best allocation plan by using the quantity and time of sanitation resource mobilization as evaluation factors and employing the TOPSIS method.

[0133] Furthermore, the method according to the embodiments of this application can also be achieved by means of... Figure 5 The architecture of the electronic device shown is used to implement this. For example... Figure 5 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, ROM 503, RAM 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as ROM 503 or hard disk 507, may store a smart sanitation management method provided in this application. The electronic device 500 may also include a user interface 508. Of course, Figure 5 The architecture shown is merely exemplary and can be omitted as needed when implementing different devices. Figure 5 One or more components in the illustrated electronic device.

[0134] Figure 6 This is a schematic diagram of a computer-readable storage medium structure provided in one embodiment of this application. Figure 6The diagram illustrates a computer-readable storage medium 600 according to one embodiment of this application. The computer-readable storage medium 600 stores computer-readable instructions. When executed by a processor, the computer-readable instructions can perform a smart sanitation management method according to an embodiment of this application, as described with reference to the above figures. The storage medium 600 includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0135] In summary, the advantages of this invention are as follows: By constructing an intelligent correlation model between dynamic events and sanitation issues, the foresight and accuracy of sanitation risk prediction are significantly improved. By systematically integrating multi-source real-time data and historical management experience, it not only accurately identifies high-probability sanitation problem types and their potential cascading risks, but also dynamically generates optimal cleaning scheduling strategies based on global constraints. Ultimately, this achieves a shift from passive response to proactive prevention, minimizing the spread and deterioration of sanitation problems while maximizing the allocation of sanitation resources, reducing overall urban sanitation operating costs, and maintaining a consistently high level of environmental sanitation quality.

[0136] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A smart sanitation management method, characterized in that, include: Several dynamic event types affecting regional sanitation conditions are identified as influencing factor entities, and the types of sanitation problems occurring in the sanitation area are identified as sanitation problem entities. A database of correlations between influencing factor entities and sanitation problem entities is constructed based on historical sanitation data. Acquire actual event data for each sub-area within the sanitation area, including at least one of pedestrian traffic, garbage generation, and abnormal weather events; Call the relational database, combine it with the current event data to predict the risk indicators of entities with health problems in each sub-region, and filter out health problem entities whose risk indicators exceed the threshold and record them as health problems to be handled; Based on sanitation management experience, determine the risk indicator intensity for each type of sanitation problem to be addressed; Based on the risk index intensity of the type of hygiene problem to be addressed, dynamically generate and implement cleaning intensity adjustment plans for corresponding sub-areas; Specifically, determining the risk indicator intensity for each type of sanitation problem to be addressed, based on sanitation management experience, includes: Based on sanitation management experience, a sanitation problem association graph is constructed, in which sanitation problem types are used as graph nodes and the inducing relationships between sanitation problems are used as directed edges; Based on historical sanitation management experience and the required processing costs for each type of sanitation problem, the intensity of the risk index for each type of sanitation problem is calculated. Based on the risk index intensity of the health problem type that will be induced by the failure to address the health problem type in a timely manner, a PAGERANK-like algorithm is used to calculate the risk index intensity of each health problem type to be addressed. The calculation process for the risk indicator intensity of the specific types of health problems to be addressed is as follows: First, the risk index intensity of all health problem types is normalized to obtain the initial PR value of all health problem types; The initial PR values ​​for all health problem types are substituted into the PAGERANK formula to obtain the analytical PR value for each health problem type. After normalizing the analytical PR values, the updated PR values ​​for each health problem type are obtained. Then, the updated PR values ​​for all health problem types are substituted into the PAGERANK formula for iterative iteration until the updated PR values ​​for all health problem types converge. Extract the convergence value of the updated PR value of the health problem type to be addressed, and use it as the risk indicator strength of the health problem type to be addressed. Specifically, the PageRank formula is: In the formula, Let be the PR value for the i-th health problem type. Let be the set of health problem types that will be induced if the i-th health problem type is not dealt with in a timely manner. for The elements within it, for The updated PR value, where N is the total number of health issue types. To cause The total number of types of health problems.

2. The intelligent sanitation management method according to claim 1, characterized in that, The database of relationships between influencing factor entities and health problem entities, constructed based on historical sanitation data, specifically includes: Extract event-problem association rules from historical sanitation data and construct a triplet knowledge base with influencing factor entities as head nodes, sanitation problem entities as tail nodes, and association relationships as edges; Based on the triplet knowledge base, a corresponding risk prediction model is established for each health problem entity. The risk prediction model adopts a logistic regression model or a neural network model.

3. The intelligent sanitation management method according to claim 2, characterized in that, When the risk prediction model is a neural network model, the process of establishing the risk prediction model is as follows: The historical event data was labeled as the training set according to the occurrence status of the health issues; The neural network is trained with event parameters as the input layer and the probability of problem occurrence as the output layer. The accuracy of the model is verified using a test set, and the model with the smallest loss function value is selected as the final risk prediction model.

4. The intelligent sanitation management method according to claim 1, characterized in that, The specific steps for calling the relational database and combining it with current event data to predict risk indicators for entities with health problems in each sub-region are as follows: For each influencing entity, sample data at multiple time scales are collected; Predict future event parameter values ​​at various scales using linear regression; Weighted fusion of multi-scale prediction values ​​yields comprehensive prediction parameters; The comprehensive prediction parameters are input into the risk prediction model, and the predicted risk index is output.

5. The intelligent sanitation management method according to claim 1, characterized in that, The process of dynamically generating and implementing a cleaning intensity adjustment plan for the corresponding sub-area based on the risk indicator intensity of the type of sanitation problem to be addressed includes: Based on sanitation experience, determine the optimal sanitation resource requirements for each sanitation problem to be addressed. Given the constraints that the sanitation resources allocated to each sanitation problem to be addressed are less than or equal to the optimal sanitation resource requirements of the sanitation problem to be addressed, and that the sum of the sanitation resources allocated to all sanitation problems to be addressed is less than the total amount of sanitation resources, several alternative sanitation allocation schemes are generated. Based on the similarity between the sanitation resources allocated to each sanitation problem in the candidate sanitation allocation scheme and the optimal sanitation resource demand of the sanitation problem, the sanitation resources are used as the treatment intensity of the sanitation problem. The treatment intensity of the sanitation problem is multiplied by the risk index intensity of the sanitation problem type to obtain the treatment index of the sanitation problem. The total problem-solving index for the candidate sanitation allocation scheme is obtained by summing up the processing indicators of all pending sanitation problems. The candidate sanitation allocation scheme with the minimum total problem handling index is selected as the optimal sanitation allocation scheme. Based on the optimal sanitation allocation plan and the current sanitation resource allocation, using the quantity and time of sanitation resource mobilization as evaluation factors, the TOPSIS method is used to screen out the optimal cleaning intensity adjustment plan that transforms the current sanitation resource allocation into the optimal sanitation allocation plan.

6. A smart sanitation management system, characterized in that, To implement the intelligent sanitation management method as described in any one of claims 1-5, comprising: The data acquisition module is used to acquire actual event data of each sub-area within the sanitation area. The event data includes at least one of the following: pedestrian flow, garbage generation, and abnormal weather events. The knowledge base construction module is used to identify several dynamic event types that affect the regional sanitation status as influencing factor entities, determine the types of sanitation problems that occur in the sanitation area as sanitation problem entities, and build a relational database between influencing factor entities and sanitation problem entities based on historical sanitation data. The risk prediction module connects the knowledge base construction module and the data acquisition module. It is used to call the relational database and combine it with the current event data to predict the risk indicators of health problem entities in each sub-region, and to screen health problem entities whose risk indicators exceed the threshold as health problems to be handled. The risk intensity calculation module, connected to the risk prediction module, is used to determine the risk index intensity of each type of sanitation problem to be addressed based on sanitation management experience. The cleaning scheduling engine, connected to the risk intensity calculation module, is used to dynamically generate and execute cleaning intensity adjustment plans for corresponding sub-areas based on the risk index intensity of the type of hygiene problem to be addressed.

7. The intelligent sanitation management system according to claim 6, characterized in that: The knowledge base construction module includes: The rule extraction unit is used to extract event-problem association rules from historical sanitation data; The triplet storage unit is used to construct a triplet knowledge base with influencing factor entities as head nodes, health problem entities as tail nodes, and relationships as edges; The model training unit is used to build a corresponding Logistic regression model or neural network model for each health problem entity as a risk prediction model. The clean scheduling engine includes: The resource demand mapping unit is used to determine the optimal sanitation resource demand for each sanitation problem to be addressed based on sanitation experience. The scheme generation unit is used to generate several candidate sanitation allocation schemes with the constraints that allocated resources ≤ optimal resource demand and total resources ≤ total sanitation volume. The processing index calculation unit is used to calculate the cumulative value of (allocated resources / optimal resource requirements) × risk index intensity in each candidate solution as the overall problem processing index; The scheme selection unit is used to select the candidate scheme with the smallest total problem handling index as the best sanitation allocation scheme. The TOPSIS scheduling unit is used to generate an optimal cleaning intensity adjustment plan from the current allocation state to the best allocation plan by using the quantity and time of sanitation resource mobilization as evaluation factors and employing the TOPSIS method.

8. A computer device, characterized in that, include: At least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the intelligent sanitation management method as described in any one of claims 1-5.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent sanitation management method according to any one of claims 1-5.

Images