A disinfection scheduling method and system based on space-time attention and multi-objective optimization

Abstract

The present application relates to the technical field of medical disinfection. A disinfection scheduling method and system based on space-time attention and multi-objective optimization are provided. The steps are as follows: real-time collection of multi-dimensional disinfection-related data in the hospital and weighted preprocessing; a space-time attention hybrid model is constructed to calculate the comprehensive score of disinfection demand in each region and period; a multi-objective weighted optimization model is established based on the score, under the constraints of equipment capacity, time period conflict and priority, the total disinfection time, equipment idle rate and average disinfection effect are optimized, and a scheduling scheme is generated; the disinfection equipment is controlled to perform the task and real-time monitoring is performed; the regional risk is dynamically evaluated according to the monitoring data, and when the risk index exceeds the threshold, the gradient disinfection remedy is triggered and the scheduling scheme is updated. The problems of poor space-time correlation and multi-objective coordination in the existing intelligent disinfection scheduling method in the complex dynamic environment of the hospital are solved.

Description

Technical Field

[0001] This invention relates to the field of medical disinfection technology, and more specifically, to a disinfection scheduling method and system based on spatiotemporal attention and multi-objective optimization. Background Technology

[0002] With the rapid development of medical informatization and intelligentization, medical institutions have placed higher demands on the accuracy, timeliness, and resource efficiency of disinfection operations. Currently, some intelligent disinfection scheduling systems are attempting to replace traditional operation modes that rely on manual experience by introducing sensor data, pathogen detection information, and personnel flow records to construct rule-based or simple machine learning-based task allocation mechanisms. These systems typically use fixed thresholds to trigger disinfection tasks or prioritize tasks based on a single dimension (such as disinfection area or pathogen concentration), which can improve the response speed of disinfection operations in some scenarios. However, they still face significant limitations in the complex and dynamic real-world hospital environment: on the one hand, the systems often ignore the spatial cross-infection correlation between different areas and fail to model the fluctuations in disinfection demand over time, resulting in a lack of linkage and temporal rationality in the prediction results; on the other hand, the scheduling process often focuses on shortening the total time or increasing the utilization rate of individual equipment, without systematically coordinating and optimizing multiple objectives such as "disinfection efficiency, resource utilization, and disinfection effect," easily leading to problems such as task backlog, equipment idleness, or disinfection blind spots, making it difficult to achieve dynamic balance and precise adaptation of disinfection resources across the entire hospital.

[0003] While existing intelligent disinfection scheduling methods have improved automation to some extent, they still have the following shortcomings: First, in the demand prediction stage, most solutions only use independent convolutional neural networks or recurrent neural networks to process pathogens or time-series data, lacking joint modeling of "region-region" correlations and "time-time" priorities. This prevents dynamic weight allocation based on spatiotemporal attention mechanisms, resulting in a single prediction dimension and delayed response to cross-regional risk transmission. Second, at the resource scheduling level, existing methods mostly employ single-objective optimization or simple rule matching, failing to establish multi-objective weighted functions and multi-layered constraint systems. This makes it impossible to coordinate total task time and equipment idle rate while ensuring disinfection effectiveness, easily leading to resource congestion or misaligned task execution order. Furthermore, existing systems typically rely on fixed thresholds for risk alarms and remedial triggers, failing to construct a dynamic risk index and gradient response mechanism that integrates multiple factors such as pathogens, patient flow, and environment. This results in rigid risk assessment and a lack of differentiation and timeliness in remedial measures. Therefore, there is an urgent need for a disinfection scheduling method that can deeply integrate spatiotemporal attention mechanisms and multi-objective collaborative optimization to improve the overall intelligence level and comprehensive operational efficiency of hospital-wide disinfection operations. Summary of the Invention

[0004] The purpose of this invention is to provide a disinfection scheduling method and system based on spatiotemporal attention and multi-objective optimization, which aims to solve the problems of poor spatiotemporal correlation and inability to coordinate and optimize multi-objectives in the complex and dynamic environment of hospitals by existing intelligent disinfection scheduling methods.

[0005] This invention is achieved through the following technical solution:

[0006] A disinfection scheduling method based on spatiotemporal attention and multi-objective optimization includes the following steps:

[0007] Real-time collection of multi-dimensional disinfection-related data within the hospital, including pathogen detection data, personnel flow data, environmental monitoring data, and diagnosis and treatment data;

[0008] We performed weighted preprocessing on multi-dimensional disinfection-related data. By introducing data importance weighting factors and outlier correction terms, we standardized the data in different dimensions to obtain preprocessed standardized data.

[0009] Based on standardized data, a spatiotemporal attention hybrid model is constructed to calculate the comprehensive score of disinfection needs for the target area at the corresponding time period;

[0010] Based on the comprehensive score of disinfection needs, a multi-objective weighted optimization model is constructed and combined with constraints to generate a disinfection task scheduling scheme. The objective function of the scheduling scheme comprehensively optimizes the total disinfection time, equipment idle rate and average disinfection effect, and satisfies equipment capacity constraints, time period conflict constraints and task priority constraints.

[0011] Based on the disinfection task scheduling scheme, the disinfection equipment is controlled to perform disinfection tasks, and the disinfection effect and environmental parameters are monitored in real time during the execution process;

[0012] Based on real-time monitoring data, dynamic risk assessment is conducted, and the dynamic risk index of each area is calculated. When the risk index exceeds the preset threshold, a tiered disinfection remediation mechanism is triggered, and the disinfection task scheduling plan is updated.

[0013] Optionally, the specific process of real-time collection of multi-dimensional disinfection-related data within the hospital is as follows:

[0014] Medical data is collected through the hospital information system, including basic patient information, diagnostic information, and treatment information.

[0015] Pathogen detection data, including bacterial culture results and virus detection results, are collected through the laboratory information system.

[0016] The personnel positioning system collects real-time data on the movement and location of medical staff and patients.

[0017] Environmental data, including airborne bacterial count, temperature and humidity, CO2 concentration, PM2.5 concentration and noise level, is collected in real time by environmental monitoring sensors deployed in each disinfection area; the environmental monitoring sensors collect data at a preset frequency.

[0018] Optionally, the specific process of performing weighted preprocessing on multi-dimensional disinfection-related data, by introducing data importance weighting factors and outlier correction terms, and standardizing the data in different dimensions to obtain preprocessed standardized data, is as follows:

[0019] The collected multi-dimensional disinfection-related data are categorized by dimension, defining the first dimension as... The first sample The original data is Statistics Total number of samples in dimensional data mean with standard deviation ;in, Corresponding pathogen detection data, population movement data, environmental monitoring data, or diagnosis and treatment data;

[0020] For different The priority of the impact of dimensional data on disinfection needs is 1. Dimensional data importance weights ;

[0021] Calculate the first Sample No. Dimensional Data with the mean The absolute value of the deviation, if If it is an extreme outlier, then it will be determined as an extreme outlier, and the first... Outlier correction factor for dimensional data Automatically adjust to 0.5, otherwise The value is 0.05;

[0022] Calculate the first Dimension offset compensation term for dimensional data Used to balance different Difference in baseline range of dimensional data, dimensional offset compensation term The average value of all sample data in the corresponding dimension is shown in the following formula:

[0023]

[0024] For the original data After standardization, the preprocessed standardized data is obtained, as shown in the following formula:

[0025]

[0026] in, This represents the standardized data after preprocessing.

[0027] Optionally, the specific process of constructing a spatiotemporal attention hybrid model based on standardized data and calculating the comprehensive score of disinfection demand for the target area at the corresponding time period is as follows:

[0028] A spatial attention mechanism is constructed, which calculates the spatial attention of any given region within the hospital based on the vectors representing regional features in the standardized data. The region and the first Spatial correlation weight between regions The spatial correlation weight The cosine similarity between the feature vectors of the two regions is used to quantify the correlation of cross-infection risk between target regions.

[0029] Construct a time-attention mechanism to determine peak periods based on historical disinfection demand data. Calculate the first Time attention weight for each time period The time attention weights With time period During peak hours The time distance is negatively correlated, which is used to characterize the priority of disinfection needs at different time periods;

[0030] The integrated calculation of disinfection requirements score is used for the first... The region in the first For each time period, combining spatial correlation weights, temporal attention weights, pathogen data features extracted by a deep learning model, and temporal pedestrian flow data features, and incorporating environmental impact coefficients and standardized environmental parameter values, a comprehensive disinfection demand score for the target area for the corresponding time period is obtained through weighted fusion calculation; this comprehensive disinfection demand score is used to characterize the area. During the period The higher the score, the higher the scheduling priority, indicating the urgency of disinfection.

[0031] Optionally, the specific process of constructing a multi-objective weighted optimization model based on the comprehensive score of disinfection needs and combining it with constraints to generate a disinfection task scheduling scheme is as follows:

[0032] Based on the comprehensive score of disinfection needs Determine the priority sequence of each disinfection task; where a higher score indicates a higher urgency for disinfection of the target area in the corresponding time period, and it will be given priority in scheduling.

[0033] A multi-objective weighted optimization model is constructed, with minimization of the objective function as the optimization direction. The objective function is composed of three weighted factors: total disinfection time, equipment idle rate, and average disinfection effect, as shown in the following formula:

[0034]

[0035] in, This represents a multi-objective weighted optimization function; The weighting coefficient for the total disinfection time target indicates the relative importance of reducing the total time during the optimization process. The total time spent on all disinfection tasks; The weighting coefficient for the equipment idle rate target indicates the relative importance of reducing the equipment idle rate during the optimization process; The total idle rate of disinfection equipment; The weighting coefficient for the average disinfection effect target indicates the relative importance of improving the average disinfection effect during the optimization process; The average disinfection effect across all areas; meets the requirements. ,and ;

[0036] Establish a three-tier constraint system, including:

[0037] Equipment capacity constraints, for any disinfection task and the allocated disinfection equipment , must meet ;in, For equipment Its disinfection capabilities; To be assigned to the device The disinfection capacity required for the disinfection task;

[0038] Time-slot conflict constraint: the execution time slots of the same disinfection equipment must not overlap between different tasks, i.e. ;in, Indicates allocation to device Task The execution period; Indicates allocation to the same device Task The execution period;

[0039] Priority order constraints, if task The overall score for disinfection needs was higher than that for the task. Then the task The execution start time should be earlier than the task. ,Right now ;in, and Representing regions During the period and region During the period The overall score for disinfection needs; and Representing tasks and tasks The actual start time of execution;

[0040] Based on the objective function and constraints, a multi-objective optimization algorithm is used to solve the problem and generate a disinfection task scheduling scheme that satisfies equipment matching, time period non-conflict, and priority order. The disinfection task scheduling scheme includes the execution equipment, start time, end time, and disinfection parameter settings for each task.

[0041] Optionally, the specific process of controlling the disinfection equipment to perform disinfection tasks based on the disinfection task scheduling scheme, and monitoring the disinfection effect and environmental parameters in real time during the execution process, is as follows:

[0042] According to the disinfection task scheduling plan, disinfection equipment is matched and task instructions are issued;

[0043] For each disinfection task, a start command is sent to the corresponding disinfection equipment according to the disinfection parameter settings, and the movement path of the disinfection equipment is optimized based on the path planning algorithm so that the disinfection tasks are executed in sequence.

[0044] During the disinfection process, the disinfection equipment collects disinfection effect data and environmental parameters in real time through built-in sensors and external environmental monitoring sensors, including real-time air colony count, temperature and humidity, CO2 concentration, disinfection coverage area and disinfection time.

[0045] The data collected in real time will be transmitted back in real time via the Internet of Things.

[0046] Optionally, the specific process of performing dynamic risk assessment and calculating the dynamic risk index for each region based on real-time monitored data is as follows:

[0047] Based on real-time monitored pathogen data, population flow data, and environmental data, pathogen risk factors, population flow risk factors, and environmental risk factors for the target area in the current time period are calculated respectively. Among them, the pathogen risk factor is determined by the ratio of real-time pathogen concentration to a preset safety threshold; the population flow risk factor is determined by the ratio of real-time population density to historical average population density; and the environmental risk factor is calculated by weighting the degree to which temperature, humidity, and CO2 concentration deviate from the suitable range.

[0048] The pathogen risk factor, human flow risk factor, and environmental risk factor are weighted and summed according to preset weights to obtain the dynamic risk index of the target area in the current period; among them, the weight of the pathogen risk factor is greater than the weight of the human flow risk factor, and the weight of the human flow risk factor is greater than the weight of the environmental risk factor.

[0049] The calculated dynamic risk index is compared with a preset risk threshold. If the dynamic risk index exceeds the preset risk threshold, a disinfection remediation mechanism is triggered. The disinfection remediation mechanism increases the disinfection intensity in a gradient manner according to the extent to which the risk index exceeds the preset risk threshold, based on an exponential enhancement formula, and generates a corresponding emergency disinfection task, which is updated in real time to the disinfection task scheduling scheme.

[0050] Based on the same inventive concept, this invention also provides a disinfection scheduling system based on spatiotemporal attention and multi-objective optimization, used to implement the aforementioned disinfection scheduling method based on spatiotemporal attention and multi-objective optimization, comprising:

[0051] The data acquisition module is used to collect multi-dimensional disinfection-related data in the hospital in real time, including pathogen detection data, personnel flow data, environmental monitoring data, and diagnosis and treatment data.

[0052] The data preprocessing module, connected to the data acquisition module, is used to perform weighted preprocessing on the multi-dimensional disinfection-related data. By introducing data importance weighting factors and outlier correction terms, the data of different dimensions is standardized, and the preprocessed standardized data is output.

[0053] The disinfection demand prediction module is connected to the data preprocessing module and is used to construct a spatiotemporal attention hybrid model based on the standardized data to calculate the comprehensive score of disinfection demand for the target area in the corresponding time period.

[0054] The task scheduling module, connected to the disinfection demand prediction module, is used to construct a multi-objective weighted optimization model based on the comprehensive score of the disinfection demand, and generate a disinfection task scheduling scheme by combining equipment capacity constraints, time period conflict constraints and priority order constraints.

[0055] An execution control module, connected to the task scheduling module, is used to control the disinfection equipment to perform disinfection tasks based on the disinfection task scheduling scheme, and to monitor the disinfection effect and environmental parameters in real time during the execution process;

[0056] The risk assessment module, connected to the execution control module, is used to calculate the dynamic risk index of each area based on real-time monitored pathogen data, personnel flow data, and environmental data.

[0057] The remediation trigger module, connected to the risk assessment module, is used to trigger a gradient disinfection remediation mechanism and update the disinfection task scheduling scheme when the dynamic risk index exceeds a preset risk threshold.

[0058] Based on the same inventive concept, the present invention also provides an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the above-described disinfection scheduling method based on spatiotemporal attention and multi-objective optimization.

[0059] The technical solution of the present invention has at least the following advantages and beneficial effects:

[0060] By constructing a spatiotemporal attention hybrid model, not only are the historical and current data of a single region analyzed, but also the spatial correlation between regions (such as the transmission of cross-infection risk) and the demand fluctuation pattern in the time dimension are jointly modeled and dynamically weighted. This enables the system to more comprehensively and accurately calculate the comprehensive score of disinfection demand in each region during a specific period, and to make advance predictions of potential risk areas and peak periods, overcoming the limitations of traditional methods in terms of single prediction dimension and delayed response.

[0061] By establishing a multi-objective weighted optimization model with total disinfection time, equipment idle rate, and average disinfection effect as the core, and comprehensively considering multiple constraints such as equipment capacity, time period conflicts, and task priority, a globally optimal or near-optimal scheduling scheme can be generated. This scheme systematically coordinates the three major objectives of efficiency, cost, and effectiveness, effectively avoiding problems such as task backlog, equipment idleness, or disinfection blind spots that are easily caused by traditional methods. It achieves dynamic and balanced allocation of disinfection resources throughout the hospital and comprehensive improvement of overall operational efficiency.

[0062] During the execution phase, a dynamic risk assessment and tiered remedial mechanism based on real-time monitoring data is introduced. By integrating multiple factors such as pathogens, population flow, and environment to calculate a dynamic risk index and setting differentiated response thresholds, it can identify areas where risks exceed limits in real time and automatically trigger matching remedial disinfection tasks. The scheduling plan is updated in a timely manner, forming a closed-loop management of "monitoring-assessment-response-optimization". This enhances the system's adaptability to complex and dynamic hospital environments and ensures the continuous reliability of disinfection effects. Attached Figure Description

[0063] Figure 1 This is a flowchart illustrating the disinfection scheduling method based on spatiotemporal attention and multi-objective optimization according to an embodiment of the present invention.

[0064] Figure 2 This is a schematic diagram of the disinfection scheduling system based on spatiotemporal attention and multi-objective optimization according to an embodiment of the present invention. Detailed Implementation

[0065] The following is a detailed description of the embodiments, in conjunction with the accompanying drawings.

[0066] Reference Figure 1A disinfection scheduling method based on spatiotemporal attention and multi-objective optimization includes the following steps:

[0067] Step 1: Collect multi-dimensional disinfection-related data in the hospital in real time. The multi-dimensional disinfection-related data includes pathogen detection data, personnel flow data, environmental monitoring data, and diagnosis and treatment data.

[0068] In some embodiments, the specific process of real-time collection of multi-dimensional disinfection-related data within the hospital is as follows:

[0069] Medical data is collected through the hospital information system, including basic patient information, diagnostic information, and treatment information.

[0070] Pathogen detection data, including bacterial culture results and virus detection results, are collected through the laboratory information system.

[0071] The personnel positioning system collects real-time data on the movement and location of medical staff and patients.

[0072] Environmental data, including airborne bacterial count, temperature and humidity, CO2 concentration, PM2.5 concentration and noise level, is collected in real time by environmental monitoring sensors deployed in each disinfection area; the environmental monitoring sensors collect data at a preset frequency.

[0073] Step 2: Perform weighted preprocessing on multi-dimensional disinfection-related data. By introducing data importance weighting factors and outlier correction terms, standardize the data in different dimensions to obtain preprocessed standardized data.

[0074] In some embodiments, the specific process of performing weighted preprocessing on multi-dimensional disinfection-related data, by introducing data importance weighting factors and outlier correction terms, and standardizing data of different dimensions to obtain preprocessed standardized data, is as follows:

[0075] The collected multi-dimensional disinfection-related data are categorized by dimension, defining the first dimension as... The first sample The original data is Statistics Total number of samples in dimensional data mean with standard deviation ;in, Corresponding pathogen detection data, population movement data, environmental monitoring data, or diagnosis and treatment data;

[0076] For different The priority of the impact of dimensional data on disinfection needs is 1. Dimensional data importance weights ;

[0077] Calculate the first Sample No. Dimensional Data with the mean The absolute value of the deviation, if If it is an extreme outlier, then it will be determined as an extreme outlier, and the first... Outlier correction factor for dimensional data Automatically adjust to 0.5, otherwise The value is 0.05;

[0078] Calculate the first Dimension offset compensation term for dimensional data Used to balance different Difference in baseline range of dimensional data, dimensional offset compensation term The average value of all sample data in the corresponding dimension is shown in the following formula:

[0079]

[0080] For the original data After standardization, the preprocessed standardized data is obtained, as shown in the following formula:

[0081]

[0082] in, This represents the standardized data after preprocessing.

[0083] Step 3: Based on standardized data, construct a spatiotemporal attention hybrid model to calculate the comprehensive score of disinfection needs for the target area at the corresponding time period.

[0084] In some embodiments, the specific process of constructing a spatiotemporal attention hybrid model based on standardized data and calculating the comprehensive score of disinfection demand for the target area in the corresponding time period is as follows:

[0085] A spatial attention mechanism is constructed, which calculates the spatial attention of any given region within the hospital based on the vectors representing regional features in the standardized data. The region and the first Spatial correlation weight between regions The spatial correlation weight The cosine similarity between the feature vectors of the two regions is used to quantify the correlation of cross-infection risk between target regions.

[0086] Construct a time-attention mechanism to determine peak periods based on historical disinfection demand data. Calculate the first Time attention weight for each time period The time attention weights With time period During peak hours The time distance is negatively correlated, which is used to characterize the priority of disinfection needs at different time periods;

[0087] The integrated calculation of disinfection requirements score is used for the first... The region in the first For each time period, combining spatial correlation weights, temporal attention weights, pathogen data features extracted by a deep learning model, and temporal pedestrian flow data features, and incorporating environmental impact coefficients and standardized environmental parameter values, a comprehensive disinfection demand score for the target area for the corresponding time period is obtained through weighted fusion calculation; this comprehensive disinfection demand score is used to characterize the area. During the period The higher the score, the higher the scheduling priority, indicating the urgency of disinfection.

[0088] Step 4: Based on the comprehensive score of disinfection needs, construct a multi-objective weighted optimization model and combine it with constraints to generate a disinfection task scheduling scheme. The objective function of the scheduling scheme comprehensively optimizes the total disinfection time, equipment idle rate and average disinfection effect, and satisfies equipment capacity constraints, time period conflict constraints and task priority constraints.

[0089] In some embodiments, the specific process of constructing a multi-objective weighted optimization model based on the comprehensive score of disinfection needs and combining it with constraints to generate a disinfection task scheduling scheme is as follows:

[0090] Based on the comprehensive score of disinfection needs Determine the priority sequence of each disinfection task; where a higher score indicates a higher urgency for disinfection of the target area in the corresponding time period, and it will be given priority in scheduling.

[0091] A multi-objective weighted optimization model is constructed, with minimization of the objective function as the optimization direction. The objective function is composed of three weighted factors: total disinfection time, equipment idle rate, and average disinfection effect, as shown in the following formula:

[0092]

[0093] in, This represents a multi-objective weighted optimization function; The weighting coefficient for the total disinfection time target indicates the relative importance of reducing the total time during the optimization process. The total time spent on all disinfection tasks; The weighting coefficient for the equipment idle rate target indicates the relative importance of reducing the equipment idle rate during the optimization process; The total idle rate of disinfection equipment; The weighting coefficient for the average disinfection effect target indicates the relative importance of improving the average disinfection effect during the optimization process; The average disinfection effect across all areas; meets the requirements. ,and ;

[0094] Establish a three-tier constraint system, including:

[0095] Equipment capacity constraints, for any disinfection task and the allocated disinfection equipment , must meet ;in, For equipment Its disinfection capabilities; To be assigned to the device The disinfection capacity required for the disinfection task;

[0096] Time-slot conflict constraint: the execution time slots of the same disinfection equipment must not overlap between different tasks, i.e. ;in, Indicates allocation to device Task The execution period; Indicates allocation to the same device Task The execution period;

[0097] Priority order constraints, if task The overall score for disinfection needs was higher than that for the task. Then the task The execution start time should be earlier than the task. ,Right now ;in, and Representing regions During the period and region During the period The overall score for disinfection needs; and Representing tasks and tasks The actual start time of execution;

[0098] Based on the objective function and constraints, a multi-objective optimization algorithm is used to solve the problem and generate a disinfection task scheduling scheme that satisfies equipment matching, time period non-conflict, and priority order. The disinfection task scheduling scheme includes the execution equipment, start time, end time, and disinfection parameter settings for each task.

[0099] Step 5: Based on the disinfection task scheduling plan, control the disinfection equipment to perform disinfection tasks, and monitor the disinfection effect and environmental parameters in real time during the execution process.

[0100] In some embodiments, the specific process of controlling the disinfection equipment to perform disinfection tasks based on the disinfection task scheduling scheme, and monitoring the disinfection effect and environmental parameters in real time during the execution process is as follows:

[0101] According to the disinfection task scheduling plan, disinfection equipment is matched and task instructions are issued;

[0102] For each disinfection task, a start command is sent to the corresponding disinfection equipment according to the disinfection parameter settings, and the movement path of the disinfection equipment is optimized based on the path planning algorithm so that the disinfection tasks are executed in sequence.

[0103] During the disinfection process, the disinfection equipment collects disinfection effect data and environmental parameters in real time through built-in sensors and external environmental monitoring sensors, including real-time air colony count, temperature and humidity, CO2 concentration, disinfection coverage area and disinfection time.

[0104] The data collected in real time will be transmitted back in real time via the Internet of Things.

[0105] Step 6: Based on real-time monitoring data, conduct dynamic risk assessment, calculate the dynamic risk index for each area, and trigger a tiered disinfection remediation mechanism when the risk index exceeds a preset threshold, and update the disinfection task scheduling plan.

[0106] In some embodiments, the specific process of performing dynamic risk assessment and calculating the dynamic risk index for each region based on real-time monitored data is as follows:

[0107] Based on real-time monitored pathogen data, population flow data, and environmental data, pathogen risk factors, population flow risk factors, and environmental risk factors for the target area in the current time period are calculated respectively. Among them, the pathogen risk factor is determined by the ratio of real-time pathogen concentration to a preset safety threshold; the population flow risk factor is determined by the ratio of real-time population density to historical average population density; and the environmental risk factor is calculated by weighting the degree to which temperature, humidity, and CO2 concentration deviate from the suitable range.

[0108] The dynamic risk index of the target area in the current period is obtained by weighting and summing the pathogen risk factors, population flow risk factors, and environmental risk factors according to preset weights; wherein, the weight of pathogen risk factors is greater than that of population flow risk factors, and the weight of population flow risk factors is greater than that of environmental risk factors. The formula for calculating the dynamic risk index is as follows:

[0109]

[0110] in, Indicates the region During the period The dynamic risk index; The pathogen risk factor is represented by the following formula:

[0111]

[0112] in, Real-time pathogen concentration; Preset safety threshold;

[0113] The risk factor for abortion is represented by the following formula:

[0114]

[0115] in, Real-time pedestrian density; The historical average population density;

[0116] The environmental risk factor is calculated by weighting parameters such as temperature, humidity, and CO2 concentration. The formula can be further expanded as shown below:

[0117]

[0118] in, For the first Real-time values ​​of environmental parameters; For the corresponding appropriate value; This is the weight of the item;

[0119] , and Let represent the weight coefficients of pathogen risk factors, human migration risk factors, and environmental risk factors, respectively, satisfying the following constraint:

[0120]

[0121] The calculated dynamic risk index is compared with a preset risk threshold. If the dynamic risk index exceeds the preset risk threshold, a disinfection remediation mechanism is triggered. This mechanism, based on the extent to which the risk index exceeds the preset risk threshold, progressively increases the disinfection intensity according to an exponential enhancement formula, and generates a corresponding emergency disinfection task, which is updated in real-time to the disinfection task scheduling scheme. When the disinfection remediation mechanism is triggered, the remediation intensity is calculated using the following formula:

[0122]

[0123] in, This indicates a preset risk threshold; This represents the floor function; Indicates the basic disinfection intensity; This indicates the adjusted intensity of remedial disinfection.

[0124] Based on the same inventive concept, and corresponding to any of the above embodiments, refer to... Figure 2 This invention provides a disinfection scheduling system based on spatiotemporal attention and multi-objective optimization, used to implement the aforementioned disinfection scheduling method based on spatiotemporal attention and multi-objective optimization, comprising:

[0125] The data acquisition module is used to collect multi-dimensional disinfection-related data in the hospital in real time, including pathogen detection data, personnel flow data, environmental monitoring data, and diagnosis and treatment data.

[0126] The data preprocessing module, connected to the data acquisition module, is used to perform weighted preprocessing on the multi-dimensional disinfection-related data. By introducing data importance weighting factors and outlier correction terms, the data of different dimensions is standardized, and the preprocessed standardized data is output.

[0127] The disinfection demand prediction module is connected to the data preprocessing module and is used to construct a spatiotemporal attention hybrid model based on the standardized data to calculate the comprehensive score of disinfection demand for the target area in the corresponding time period.

[0128] The task scheduling module, connected to the disinfection demand prediction module, is used to construct a multi-objective weighted optimization model based on the comprehensive score of the disinfection demand, and generate a disinfection task scheduling scheme by combining equipment capacity constraints, time period conflict constraints and priority order constraints.

[0129] An execution control module, connected to the task scheduling module, is used to control the disinfection equipment to perform disinfection tasks based on the disinfection task scheduling scheme, and to monitor the disinfection effect and environmental parameters in real time during the execution process;

[0130] The risk assessment module, connected to the execution control module, is used to calculate the dynamic risk index of each area based on real-time monitored pathogen data, personnel flow data, and environmental data.

[0131] The remediation trigger module, connected to the risk assessment module, is used to trigger a gradient disinfection remediation mechanism and update the disinfection task scheduling scheme when the dynamic risk index exceeds a preset risk threshold.

[0132] Based on the same inventive concept, corresponding to any of the above embodiments, the present invention provides an electronic device, including a memory and a processor. The memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the disinfection scheduling method based on spatiotemporal attention and multi-objective optimization of the embodiments.

[0133] Alternatively, the aforementioned electronic device may be a server.

[0134] It is understood that the processor in the embodiments of the present invention may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0135] The method steps in the embodiments of the present invention can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art.

Claims

1. A disinfection scheduling method based on spatiotemporal attention and multi-objective optimization, characterized in that, Includes the following steps: Real-time collection of multi-dimensional disinfection-related data within the hospital, including pathogen detection data, personnel flow data, environmental monitoring data, and diagnosis and treatment data; We performed weighted preprocessing on multi-dimensional disinfection-related data. By introducing data importance weighting factors and outlier correction terms, we standardized the data in different dimensions to obtain preprocessed standardized data. Based on standardized data, a spatiotemporal attention hybrid model is constructed to calculate the comprehensive score of disinfection demand for the target area during the corresponding time period. The specific process of constructing the spatiotemporal attention hybrid model based on standardized data and calculating the comprehensive score of disinfection demand for the target area during the corresponding time period is as follows: A spatial attention mechanism is constructed, which calculates the spatial attention of any given region within the hospital based on the vectors representing regional features in the standardized data. The region and the first Spatial correlation weight between regions The spatial correlation weight The cosine similarity between the feature vectors of the two regions is used to quantify the correlation of cross-infection risk between target regions. Construct a time-attention mechanism to determine peak periods based on historical disinfection demand data. Calculate the first Time attention weight for each time period The time attention weights With time period During peak hours The time distance is negatively correlated, which is used to characterize the priority of disinfection needs at different time periods; The integrated calculation of disinfection requirements score is used for the first... The region in the first For each time period, combining spatial correlation weights, temporal attention weights, pathogen data features extracted by a deep learning model, and temporal pedestrian flow data features, and incorporating environmental impact coefficients and standardized environmental parameter values, a comprehensive disinfection demand score for the target area for the corresponding time period is obtained through weighted fusion calculation; this comprehensive disinfection demand score is used to characterize the area. During the period The higher the score, the higher the priority of the disinfection; Based on the comprehensive score of disinfection needs, a multi-objective weighted optimization model is constructed and combined with constraints to generate a disinfection task scheduling scheme. The objective function of the scheduling scheme comprehensively optimizes the total disinfection time, equipment idle rate and average disinfection effect, and satisfies equipment capacity constraints, time period conflict constraints and task priority constraints. Based on the disinfection task scheduling scheme, the disinfection equipment is controlled to perform disinfection tasks, and the disinfection effect and environmental parameters are monitored in real time during the execution process; Based on real-time monitoring data, dynamic risk assessment is conducted, and the dynamic risk index of each area is calculated. When the risk index exceeds the preset threshold, a tiered disinfection remediation mechanism is triggered, and the disinfection task scheduling plan is updated.

2. The disinfection scheduling method based on spatiotemporal attention and multi-objective optimization as described in claim 1, characterized in that, The specific process of real-time collection of multi-dimensional disinfection-related data within the hospital is as follows: Medical data is collected through the hospital information system, including basic patient information, diagnostic information, and treatment information. Pathogen detection data, including bacterial culture results and virus detection results, are collected through the laboratory information system. The personnel positioning system collects real-time data on the movement and location of medical staff and patients. Environmental data, including airborne bacterial count, temperature and humidity, CO2 concentration, PM2.5 concentration and noise level, is collected in real time by environmental monitoring sensors deployed in each disinfection area; the environmental monitoring sensors collect data at a preset frequency.

3. The disinfection scheduling method based on spatiotemporal attention and multi-objective optimization as described in claim 1, characterized in that, The specific process of performing weighted preprocessing on multi-dimensional disinfection-related data, by introducing data importance weighting factors and outlier correction terms, and standardizing the data in different dimensions to obtain preprocessed standardized data, is as follows: The collected multi-dimensional disinfection-related data are categorized by dimension, defining the first dimension as... The first sample The original data is Statistics Total number of samples in dimensional data mean with standard deviation ;in, Corresponding pathogen detection data, population movement data, environmental monitoring data, or diagnosis and treatment data; For different The priority of the impact of dimensional data on disinfection needs is 1. Dimensional data importance weights ; Calculate the first Sample No. Dimensional Data with the mean The absolute value of the deviation, if If it is an extreme outlier, then it will be determined as an extreme outlier, and the first... Outlier correction factor for dimensional data Automatically adjust to 0.5, otherwise The value is 0.05; Calculate the first Dimension offset compensation term for dimensional data Used to balance different Difference in baseline range of dimensional data, dimensional offset compensation term The average value of all sample data in the corresponding dimension is shown in the following formula: For the original data After standardization, the preprocessed standardized data is obtained, as shown in the following formula: in, This represents the standardized data after preprocessing.

4. The disinfection scheduling method based on spatiotemporal attention and multi-objective optimization as described in claim 1, characterized in that, The specific process of constructing a multi-objective weighted optimization model based on the comprehensive score of disinfection needs and combining it with constraints to generate a disinfection task scheduling scheme is as follows: Based on the comprehensive score of disinfection needs Determine the priority sequence of each disinfection task; where a higher score indicates a higher urgency for disinfection of the target area in the corresponding time period, and it will be given priority in scheduling. A multi-objective weighted optimization model is constructed, with minimization of the objective function as the optimization direction. The objective function is composed of three weighted factors: total disinfection time, equipment idle rate, and average disinfection effect, as shown in the following formula: in, This represents a multi-objective weighted optimization function; The weighting coefficient for the total disinfection time target indicates the relative importance of reducing the total time during the optimization process. The total time spent on all disinfection tasks; The weighting coefficient for the equipment idle rate target indicates the relative importance of reducing the equipment idle rate during the optimization process; The total idle rate of disinfection equipment; The weighting coefficient for the average disinfection effect target indicates the relative importance of improving the average disinfection effect during the optimization process; The average disinfection effect across all areas; meets the requirements. ,and ; Establish a three-tier constraint system, including: Equipment capacity constraints, for any disinfection task and the allocated disinfection equipment , must meet ;in, For equipment Its disinfection capabilities; To be assigned to the device The disinfection capacity required for the disinfection task; Time-slot conflict constraint: the execution time slots of the same disinfection equipment must not overlap between different tasks, i.e. ;in, Indicates allocation to device Task The execution period; Indicates allocation to the same device Task The execution period; Priority order constraints, if task The overall score for disinfection needs was higher than that for the task. Then the task The execution start time should be earlier than the task. ,Right now ;in, and Representing regions During the period and region During the period The overall score for disinfection needs; and Representing tasks and tasks The actual start time of execution; Based on the objective function and constraints, a multi-objective optimization algorithm is used to solve the problem and generate a disinfection task scheduling scheme that satisfies equipment matching, time period non-conflict, and priority order. The disinfection task scheduling scheme includes the execution equipment, start time, end time, and disinfection parameter settings for each task.

5. The disinfection scheduling method based on spatiotemporal attention and multi-objective optimization as described in claim 1, characterized in that, The specific process of controlling disinfection equipment to perform disinfection tasks based on the disinfection task scheduling scheme, and monitoring the disinfection effect and environmental parameters in real time during the execution process is as follows: According to the disinfection task scheduling plan, disinfection equipment is matched and task instructions are issued; For each disinfection task, a start command is sent to the corresponding disinfection equipment according to the disinfection parameter settings, and the movement path of the disinfection equipment is optimized based on the path planning algorithm so that the disinfection tasks are executed in sequence. During the disinfection process, the disinfection equipment collects disinfection effect data and environmental parameters in real time through built-in sensors and external environmental monitoring sensors, including real-time air colony count, temperature and humidity, CO2 concentration, disinfection coverage area and disinfection time. The data collected in real time will be transmitted back in real time via the Internet of Things.

6. The disinfection scheduling method based on spatiotemporal attention and multi-objective optimization as described in claim 5, characterized in that, The specific process of conducting dynamic risk assessment and calculating the dynamic risk index for each region based on real-time monitoring data is as follows: Based on real-time monitored pathogen data, population flow data, and environmental data, pathogen risk factors, population flow risk factors, and environmental risk factors for the target area in the current time period are calculated respectively. Among them, the pathogen risk factor is determined by the ratio of real-time pathogen concentration to a preset safety threshold; the population flow risk factor is determined by the ratio of real-time population density to historical average population density; and the environmental risk factor is calculated by weighting the degree to which temperature, humidity, and CO2 concentration deviate from the suitable range. The pathogen risk factor, human flow risk factor, and environmental risk factor are weighted and summed according to preset weights to obtain the dynamic risk index of the target area in the current period; among them, the weight of the pathogen risk factor is greater than the weight of the human flow risk factor, and the weight of the human flow risk factor is greater than the weight of the environmental risk factor. The calculated dynamic risk index is compared with a preset risk threshold. If the dynamic risk index exceeds the preset risk threshold, a disinfection remediation mechanism is triggered. The disinfection remediation mechanism increases the disinfection intensity in a gradient manner according to the extent to which the risk index exceeds the preset risk threshold, based on an exponential enhancement formula, and generates a corresponding emergency disinfection task, which is updated in real time to the disinfection task scheduling scheme.

7. A disinfection scheduling system based on spatiotemporal attention and multi-objective optimization, used to implement the disinfection scheduling method based on spatiotemporal attention and multi-objective optimization as described in any one of claims 1-6, characterized in that, include: The data acquisition module is used to collect multi-dimensional disinfection-related data in the hospital in real time, including pathogen detection data, personnel flow data, environmental monitoring data, and diagnosis and treatment data. The data preprocessing module, connected to the data acquisition module, is used to perform weighted preprocessing on the multi-dimensional disinfection-related data. By introducing data importance weighting factors and outlier correction terms, the data of different dimensions is standardized, and the preprocessed standardized data is output. The disinfection demand prediction module is connected to the data preprocessing module and is used to construct a spatiotemporal attention hybrid model based on the standardized data to calculate the comprehensive score of disinfection demand for the target area in the corresponding time period. The task scheduling module, connected to the disinfection demand prediction module, is used to construct a multi-objective weighted optimization model based on the comprehensive score of the disinfection demand, and generate a disinfection task scheduling scheme by combining equipment capacity constraints, time period conflict constraints and priority order constraints. An execution control module, connected to the task scheduling module, is used to control the disinfection equipment to perform disinfection tasks based on the disinfection task scheduling scheme, and to monitor the disinfection effect and environmental parameters in real time during the execution process; The risk assessment module, connected to the execution control module, is used to calculate the dynamic risk index of each area based on real-time monitored pathogen data, personnel flow data, and environmental data. The remediation trigger module, connected to the risk assessment module, is used to trigger a gradient disinfection remediation mechanism and update the disinfection task scheduling scheme when the dynamic risk index exceeds a preset risk threshold.

8. An electronic device, characterized in that, The device includes a memory and a processor, the memory being used to store a computer program, and the processor running the computer program to cause the electronic device to perform the disinfection scheduling method based on spatiotemporal attention and multi-objective optimization as described in any one of claims 1-6.

Images