Control method and system of intelligent food sample reserving cabinet

Abstract

The invention discloses a control method and system for an intelligent food sample reserving cabinet, and relates to the technical field of resource allocation, and the method comprises the steps: obtaining a historical resource learning sample and a total resource amount of the intelligent food sample reserving cabinet, and generating a prediction task scheduling mode; obtaining a current use scene of the intelligent food sample reserving cabinet, and adjusting the prediction task scheduling mode to obtain a target task scheduling mode; based on the target task scheduling mode, the dynamic weight of each task is obtained, the corresponding task is marked as a burst task, and resource preemption is automatically carried out on the burst task; monitoring a system of the intelligent food sample reserving cabinet to obtain abnormal data, and dynamically regulating and controlling the task according to the abnormal data and the dynamic weight; and performing real-time processing on the real-time task according to the target task scheduling mode, and uploading the delayed task to the cloud. The method has the effect of improving the rationality and timeliness of the intelligent food sample reserving cabinet in resource allocation.

Description

Technical Field

[0001] The present application relates to the technical field of resource allocation, and in particular to a control method and system for an intelligent food sample storage cabinet. Background Art

[0002] The main function of the intelligent food sample cabinet is to store video samples, ensure the standardization and safety of food samples, prevent video contamination and deterioration, and ensure dietary health.

[0003] In the prior art, smart food sample cabinets use various sensors to detect key parameters of food such as temperature and humidity in real time during use, and issue alarms when anomalies occur. In actual application, smart food sample cabinets need to process various data tasks in the cabinet, such as temperature monitoring, data recording, communication upload, etc. These tasks will simultaneously seize computing resources, causing system delays or freezes. At the same time, when faced with sudden emergency tasks, the response is delayed due to the resource occupation of routine tasks, which in turn causes the food in the sample cabinet to not be processed in a timely manner. Therefore, the resource scheduling and allocation of smart food sample cabinets has become an urgent problem to be solved. Summary of the Invention

[0004] The purpose of the present invention is to provide a control method and system for an intelligent food sample storage cabinet to solve the problems raised in the above background technology.

[0005] In a first aspect, the present application provides a method for controlling an intelligent food sample storage cabinet, the method comprising: Obtain historical resource learning samples and total resource amounts of the smart food sample cabinet, obtain a resource allocation curve based on the historical resource learning samples and the total resource amounts, and generate a prediction task scheduling mode for the current smart food sample cabinet according to the resource allocation curve; Obtaining a current usage scenario of the smart food sample cabinet, adjusting the predicted task scheduling mode according to the usage scenario, and obtaining a target task scheduling mode; Based on the target task scheduling mode, a dynamic weight of each task is obtained, and the dynamic weight is monitored. When the dynamic weight is higher than a preset weight threshold, the corresponding task is marked as a burst task, and the burst task automatically preempts resources; Monitor the system of the intelligent food sample cabinet to obtain abnormal data, and dynamically adjust the task according to the abnormal data and the dynamic weight; According to the target task scheduling mode, the task data of the smart food sample cabinet is collected, the task data is classified into real-time tasks and delayed tasks, the real-time tasks are processed in real time, and the delayed tasks are uploaded to the cloud.

[0006] Preferably, the steps of obtaining historical resource learning samples of the smart food sample cabinet, obtaining a resource allocation curve based on the historical resource learning samples and the total amount of resources, and generating a prediction task scheduling mode of the current smart food sample cabinet according to the resource allocation curve are specifically as follows: Obtain historical resource learning samples of the smart food sample cabinet, construct a timeline, and expand the historical resource learning samples on the timeline to obtain a sample data change axis; According to the sample data change axis, a historical resource change law and a historical resource change value are obtained, and a historical resource change curve is generated according to the historical resource change law and the historical resource change value; Obtaining the total amount of resources, obtaining a historical resource occupancy curve based on the total amount of resources and the historical resource change curve, and generating a resource allocation curve based on the historical resource occupancy curve; The total number of tasks of the intelligent food sample retention cabinet is obtained, and a forecast task scheduling model is generated by combining the resource allocation curve and the total number of tasks.

[0007] Preferably, the steps of obtaining the total number of tasks of the intelligent food sample cabinet and generating a predicted task scheduling model in combination with the resource allocation curve and the total number of tasks are specifically as follows: Obtain the total number of tasks for the smart food sample cabinet and extract the amount of task resources required for the execution of each task; According to the resource allocation curve, the resource allocation curve is used to identify peak resources to obtain peak time periods and peak tasks corresponding to the peak time periods; Extracting the task type and resource occupancy of the peak task, and generating resource redundancy according to the task type and resource occupancy; Extracting the task importance value of each task based on the peak period to obtain idle tasks, and obtaining the idle resource amount according to the idle tasks and the task resource amount; The idle resource amount is scheduled and predicted according to the peak period, the peak task and the resource redundancy, and a predicted task scheduling mode is generated.

[0008] Preferably, the steps of obtaining the current usage scenario of the smart food sample cabinet, adjusting the predicted task scheduling mode according to the usage scenario, and obtaining the target task scheduling mode are specifically as follows: Obtain the number of samples in the current smart food sample cabinet, the number of cabinet door openings and closings, the cabinet ambient temperature, and network status parameters; Obtaining usage of the smart food sample cabinet according to the number of samples in the cabinet, the number of cabinet door openings and closings, and the ambient temperature in the cabinet; According to the network status parameters, the network environment of the smart food sample cabinet is obtained, and the usage scenario of the smart food sample cabinet is obtained in combination with the usage situation and the network environment; Based on the usage scenario, the task resource requirements of the smart food sample retention cabinet are obtained, and the predicted task scheduling mode is adjusted according to the task resource requirements to obtain the target task scheduling mode.

[0009] Preferably, the step of obtaining the dynamic weight of each task based on the target task scheduling mode is specifically as follows: Based on the target task scheduling mode, the work content type of each task in the working process of the intelligent food sample cabinet is obtained; Based on the work content type, each task is evaluated for importance to obtain an importance value of each task, and an initial weight value is obtained according to the importance value; extracting the peak task type of each of the peak tasks and the peak period length of each of the peak periods based on the resource allocation curve; The initial weight values are dynamically programmed according to the peak task type and the peak period length to obtain a dynamic weight for each task.

[0010] Preferably, the dynamic weight is monitored, and when the dynamic weight is higher than a preset weight threshold, the corresponding task is marked as a burst task, and the burst task automatically preempts resources, specifically: Monitoring the dynamic weights to obtain a current weight value of each dynamic weight, and determining whether the current weight value is higher than a preset weight threshold; If it is determined that the current weight value is higher than the weight threshold, marking the current weight value as a burst weight value, and marking the task corresponding to the burst weight value as a burst task; Determine whether the current weight value is lower than or equal to a preset weight lower limit threshold; if it is determined that the current weight value is lower than the weight lower limit threshold, mark the current weight value as a buffer weight value, and mark the task corresponding to the buffer weight value as a buffer task; The buffer task is closed to release the buffer resources of the buffer task, and the buffer resources are added to the burst task.

[0011] Preferably, the system of the intelligent food sample cabinet is monitored to obtain abnormal data, and the step of dynamically regulating the task according to the abnormal data and the dynamic weight is specifically as follows: Monitor the system of the smart food sample cabinet to obtain hardware temperature data and software operation data; Based on the hardware temperature data, the current working efficiency of the system hardware is obtained, and based on the software operation data, the current fluency of the system software is obtained; Based on the current working efficiency and the current fluency, performing an abnormality assessment on the hardware temperature data and the software operation data to obtain abnormal data, and obtaining the data processing capability of the system according to the abnormal data; According to the task content type, the tasks are divided into basic tasks, core tasks and supplementary tasks; Obtaining the task priorities of the core tasks, and sorting them according to the task priorities to generate a core task sequence; According to the data processing capabilities, the basic tasks are effectively guaranteed, and according to the core task sequence, priority is given to guarantee, and the additional tasks are shut down.

[0012] Preferably, the steps of collecting task data of the smart food sample cabinet, classifying the task data to obtain real-time tasks and deferred tasks, processing the real-time tasks in real time, and uploading the deferred tasks to the cloud are specifically as follows: Collecting task data of the intelligent food sample cabinet, and classifying the task data into real-time tasks and deferred tasks; Based on the real-time task, calling the system of the intelligent food sample cabinet to process the real-time task in real time; Based on the deferred task, uploading the deferred task to the cloud, calling the cloud to process the deferred task, and generating a return data packet; The busy state of the system is acquired, the idle time of the busy state is collected to obtain an idle time period, and the return data packet is sent back to the smart food sample cabinet within the idle time period.

[0013] In a second aspect, the present application provides a control system for an intelligent food sample storage cabinet, the system comprising: Scheduling prediction module: used to obtain historical resource learning samples and total resource amount of the smart food sample cabinet, obtain a resource allocation curve based on the historical resource learning samples and the total resource amount, and generate a prediction task scheduling mode for the current smart food sample cabinet according to the resource allocation curve; Scheduling adjustment module: used to obtain the current usage scenario of the smart food sample cabinet, adjust the predicted task scheduling mode according to the usage scenario, and obtain the target task scheduling mode; Resource preemption module: used to obtain the dynamic weight of each task based on the target task scheduling mode, and monitor the dynamic weight. When the dynamic weight is higher than the preset weight threshold, the corresponding task is marked as a burst task, and the burst task automatically preempts resources; Abnormal allocation module: used to monitor the system of the intelligent food sample cabinet, obtain abnormal data, and dynamically adjust the tasks according to the abnormal data and the dynamic weight; Data processing module: used to collect task data of the smart food sample cabinet according to the target task scheduling mode, classify the task data into real-time tasks and delayed tasks, process the real-time tasks in real time, and upload the delayed tasks to the cloud.

[0014] In summary, this application includes at least one of the following beneficial technical effects: By obtaining historical resource learning samples and total resource volume from the smart food sample cabinet, a prediction task scheduling model is generated. The current usage scenario of the smart food sample cabinet is then determined and the prediction task scheduling model is adjusted based on this scenario to obtain the target task scheduling model. The target task scheduling model then determines the dynamic weight of each task. This dynamic weight changes as the usage scenario evolves. When the dynamic weight exceeds the weight threshold, the corresponding task is marked as a burst task. This burst task retrieves the current weights of other tasks and preempts resources from tasks with lower weights to ensure sufficient resources. The smart food sample cabinet system is then monitored to obtain abnormal data. Based on this abnormal data and the dynamic weights, tasks are dynamically adjusted to ensure that the most basic tasks of the smart food safety cabinet are executed normally and that non-basic tasks are executed as needed. Finally, based on the target task scheduling model, task data from the smart food sample cabinet is collected and classified into real-time and deferred tasks. The system processes the real-time tasks in real time to ensure timely processing. Deferred tasks are then sent to the cloud and processed by the cloud to ensure system smoothness. Through the above methods, the rationality and timeliness of resource allocation of smart food sample cabinets are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a flowchart of the steps of a control method of an intelligent food sample storage cabinet provided in an embodiment of the present application; Figure 2 This is a module block diagram of a control system of an intelligent food sample retention cabinet provided in an embodiment of the present application.

[0016] Explanation of the accompanying symbols: 1. Scheduling prediction module; 2. Scheduling adjustment module; 3. Resource preemption module; 4. Abnormal allocation module; 5. Data processing module. DETAILED DESCRIPTION

[0017] The following is combined with Figure 1-Figure 2 This application is further described in detail, but the embodiments of the present invention are not limited thereto.

[0018] The embodiments of the present application disclose a control method and system for an intelligent food sample storage cabinet.

[0019] In this embodiment, a method for controlling an intelligent food sample storage cabinet includes: S100: Obtain historical resource learning samples and total resources of the smart food sample cabinet, obtain a resource allocation curve based on the historical resource learning samples and total resources, and generate a prediction task scheduling mode for the current smart food sample cabinet according to the resource allocation curve; S200: Obtaining a current usage scenario of the smart food sample storage cabinet, adjusting the prediction task scheduling mode according to the usage scenario, and obtaining a target task scheduling mode; S300: Based on the target task scheduling mode, a dynamic weight of each task is obtained and monitored. When the dynamic weight is higher than a preset weight threshold, the corresponding task is marked as a burst task, and resources are automatically preempted for the burst task. S400: Monitor the system of the intelligent food sample cabinet to obtain abnormal data, and dynamically adjust the task based on the abnormal data and dynamic weights; S500: According to the target task scheduling mode, the task data of the smart food sample cabinet is collected, the task data is classified into real-time tasks and delayed tasks, the real-time tasks are processed in real time, and the delayed tasks are uploaded to the cloud.

[0020] It should be pointed out that the above modules are only the basic steps of this embodiment. During the specific implementation process, some steps can be appropriately added, reduced or modified without affecting the overall implementation effect.

[0021] The steps of obtaining historical resource learning samples of the smart food sample cabinet, obtaining a resource allocation curve based on the historical resource learning samples and the total amount of resources, and generating a prediction task scheduling mode for the current smart food sample cabinet according to the resource allocation curve are as follows: Obtain historical resource learning samples from the smart food sample cabinet, construct a timeline, and expand the historical resource learning samples on the timeline to obtain a sample data change axis; According to the sample data change axis, the historical resource change law and the historical resource change value are obtained, and the historical resource change curve is generated according to the historical resource change law and the historical resource change value; Obtain the total amount of resources, obtain the historical resource occupancy curve based on the total amount of resources and the historical resource change curve, and generate the resource allocation curve based on the historical resource occupancy curve; Obtain the total number of tasks for the smart food sample cabinet, and generate a predictive task scheduling model based on the resource allocation curve and the total number of tasks.

[0022] In practice, a hospital used a smart food sample cabinet for two years. Historical resource learning samples were obtained for this cabinet. These samples included CPU usage and memory usage data for different time periods each day over the past two years. A timeline was constructed, expanding the historical resource learning samples in chronological order, with the timeline divided into hourly units. For example, each day was divided into 24 time periods, with CPU usage recorded for each period (e.g., 60% CPU usage at 8 AM and 85% at 3 PM). This generated a sample data variation axis. This axis revealed that CPU usage consistently exceeded 80% from 10 AM to 12 PM, and memory usage exceeded 90% from 2 PM to 4 PM. The historical resource variation pattern was summarized as "peak computing from 10 AM to 12 AM, peak storage from 2 PM to 4 PM." A historical resource variation curve was generated based on these patterns. The curve shows that CPU usage rapidly increased from 60% to 85% at 10 AM, and memory usage jumped from 70% to 92% at 2 PM. The total resources of the cabinet were a quad-core 3.0 GHz CPU and 8 GB of memory. Comparing the historical resource usage curve with the total resource count revealed that CPU usage reached 85% during the morning peak (three of the four cores were fully loaded), and memory usage reached 92% in the afternoon (7.36GB of the 8GB were occupied). This generated a historical resource usage curve. Considering the total number of tasks (120 temperature monitoring tasks, 80 data upload tasks, and 20 alarm tasks per day), the predicted task scheduling pattern was "prioritizing CPU resources for temperature monitoring in the morning and memory for data upload in the afternoon."

[0023] The steps for obtaining the total number of tasks for the smart food sample cabinet and generating a predictive task scheduling model based on the resource allocation curve and the total number of tasks are as follows: Obtain the total number of tasks for the smart food sample cabinet and extract the amount of task resources required for each task during execution; According to the resource allocation curve, identify the peak resources of the resource allocation curve to obtain the peak time period and the peak tasks corresponding to the peak time period; Extract the task type and resource usage of peak tasks, and generate resource redundancy based on the task type and resource usage; Based on the peak period, the task importance value of each task is extracted to obtain the idle tasks. Based on the idle tasks and the task resource amount, the idle resource amount is obtained. The idle resource amount is scheduled and predicted based on the peak period, peak task and resource redundancy, and a predicted task scheduling model is generated.

[0024] In practice, using a smart food sample cabinet in a hospital that has been in use for two years as an example, the total number of acquisition tasks includes temperature monitoring (once every 10 minutes), data upload (once every hour), and alarm processing (triggered in real time). Each temperature monitoring task requires 15% of CPU resources and 5MB of memory; data upload requires 10% of CPU resources and 50MB of memory; and alarm processing requires 30% of CPU resources and 20MB of memory. Based on the resource allocation curve, 9:00 AM to 11:00 AM is the peak CPU period (average usage of 82%), corresponding to temperature monitoring. Memory peaks from 1:00 PM to 3:00 PM (average usage of 85%), corresponding to data upload. The temperature monitoring task type is "periodic computing task," with peak CPU usage of 6 times per hour x 15% = 90%; the data upload task type is "batch storage task," with memory usage of 1 time per hour x 50MB = 50MB. Resource redundancy is generated as 10% CPU and 15% memory. Between 9:00 AM and 11:00 AM, the data upload task's importance value was only 3 (out of 10), marking it as idle. Its hourly resource usage was 10% of the CPU (6 times = 60%) and 50 MB of memory. Due to peak demand, the data upload task was postponed to the afternoon, freeing up 60% of the morning CPU resources for temperature monitoring. This generated a predictive scheduling pattern: "Pause data uploads between 9:00 AM and 11:00 AM, allocating all CPU resources to temperature monitoring."

[0025] Obtain the current usage scenario of the smart food sample cabinet, adjust the prediction task scheduling mode according to the usage scenario, and obtain the target task scheduling mode. The specific steps are as follows: Obtain the number of samples in the current smart food sample cabinet, the number of cabinet door openings and closings, the cabinet ambient temperature, and network status parameters; The usage of the smart food sample cabinet is obtained based on the number of samples in the cabinet, the number of cabinet door openings and closings, and the ambient temperature in the cabinet; According to the network status parameters, the network environment of the smart food sample cabinet is obtained, and the usage scenario of the smart food sample cabinet is obtained by combining the usage and network environment conditions; Based on the usage scenario, the task resource requirements of the smart food sample cabinet are obtained, and the predicted task scheduling mode is adjusted according to the task resource requirements to obtain the target task scheduling mode.

[0026] In practice, a hospital's two-year-old smart food sample cabinet was used as an example. The cabinet currently held 35 samples (maximum capacity: 50), the cabinet door opened and closed an average of 8 times per hour (reaching 15 times during the morning rush hour), and the ambient temperature inside the cabinet was 6°C (the default setting was 5°C). Network status parameters indicated a Wi-Fi signal strength of -70dBm (critically weak signal state). Analysis revealed that data storage tasks increased when the number of samples reached 70% of capacity; frequent door openings and closings increased the processing load for temperature fluctuations; and weak network signals affected data upload speeds. Combined with the ambient temperature exceeding the standard by 0.5°C, the usage scenario was identified as "high-load operation." Based on this scenario, task resource requirements were adjusted as follows: the temperature monitoring task frequency was increased from every 10 minutes to every 5 minutes; the data upload task was modified from every hour to every 30 minutes; and the alarm task priority was increased to the highest. Adjust the predictive scheduling mode, cancel the video backup task originally scheduled for the afternoon, change the CPU resource allocation ratio from "temperature monitoring: data upload = 7:3" to "temperature monitoring: alarm processing = 8:2", and generate the final target task scheduling mode.

[0027] Based on the target task scheduling mode, the steps to obtain the dynamic weight of each task are as follows: Based on the target task scheduling mode, the work content type of each task in the working process of the intelligent food sample cabinet is obtained; Based on the work content type, evaluate the importance of each task, obtain the importance value of each task, and obtain the initial weight value based on the importance value; Based on the resource allocation curve, the peak task type of each peak task and the peak period length of each peak period are extracted; The initial weight values are dynamically programmed according to the peak task type and the length of the peak period to obtain the dynamic weight of each task.

[0028] In practice, for example, a smart food sample storage cabinet, used for two years at a hospital, was used. The target task scheduling model included temperature monitoring (a basic task), data upload (an auxiliary function), and alarm processing (a core task). Temperature monitoring was assigned an importance rating of 8 / 10 (basic function); data upload was assigned an importance rating of 5 / 10; and alarm processing was assigned an importance rating of 10 / 10. Initial weights were set at 0.4 for temperature monitoring, 0.3 for data upload, and 0.3 for alarm processing. Based on the resource allocation curve, the peak task type for the hours of 9:00 AM to 11:00 AM was temperature monitoring, with a duration of three hours; and the peak task for the hours of 1:00 PM to 3:00 PM was data upload, with a duration of two hours. Dynamic planning was used to adjust the weights: at 9:00 AM, the alarm processing weight increased from 0.3 to 0.5, while the data upload weight decreased from 0.3 to 0.1. At 2:00 PM, the data upload weight increased to 0.6, while the temperature monitoring weight decreased to 0.2. For example, at 10:00 AM, if the temperature was abnormal, the dynamic weight of alarm processing jumped to 0.8 in real time, while the data upload weight returned to zero, ensuring that the alarm task received 80% of the CPU resources.

[0029] Monitor the dynamic weight. When the dynamic weight is higher than the preset weight threshold, the corresponding task is marked as a burst task. The burst task automatically preempts resources. Specifically, the following steps are performed: Monitor the dynamic weights, obtain the current weight value of each dynamic weight, and determine whether the current weight value is higher than the preset weight threshold; If it is determined that the current weight value is higher than the weight threshold, the current weight value is marked as a burst weight value, and the task corresponding to the burst weight value is marked as a burst task; Determine whether the current weight value is lower than or equal to a preset weight lower limit threshold. If it is determined that the current weight value is lower than the weight lower limit threshold, mark the current weight value as a buffer weight value, and mark the task corresponding to the buffer weight value as a buffer task; Close the buffer task, release the buffer resources of the buffer task, and replenish the buffer resources for the burst task.

[0030] In practice, a smart food sample cabinet, used for two years at a hospital, was used as an example. The weight threshold was set at 0.7, and the lower weight threshold was set at 0.2. When a sudden temperature rise triggered an alarm task, the system detected that the dynamic weight of the alarm processing task reached 0.85 (exceeding the 0.7 threshold) in real time, immediately marking it as an urgent task. Simultaneously, the system detected that the weight of the data upload task had dropped to 0.15 (below 0.2), marking it as a buffer task. The system then shut down the data upload task, freeing up 15% of the CPU and 50MB of memory. The freed 15% of CPU resources was transferred entirely to the alarm task, increasing its CPU usage from 65% to 80%. The 50MB of memory was converted into cache space for storing temperature anomaly logs. The task list now became: alarm processing (weight 0.85, occupies 80% of the CPU and 70MB of memory), temperature monitoring (weight 0.35, occupies 15% of the CPU and 20MB of memory), and the remaining 5% of the CPU and 10MB of memory were reserved for system backup.

[0031] Monitor the system of the smart food sample cabinet to obtain abnormal data. Based on the abnormal data and dynamic weights, dynamically adjust the tasks. The specific steps are as follows: Monitor the system of the smart food sample cabinet to obtain hardware temperature data and software operation data; Based on the hardware temperature data, the current working efficiency of the system hardware is obtained, and based on the software running data, the current fluency of the system software is obtained; Based on the current work efficiency and current fluency, perform abnormal evaluation on hardware temperature data and software operation data to obtain abnormal data, and then obtain the data processing capability of the system based on the abnormal data; According to the task content type, tasks are divided into basic tasks, core tasks and supplementary tasks; Obtain the task priorities of core tasks, sort them according to the task priorities, and generate a core task sequence; Based on data processing capabilities, basic tasks are effectively guaranteed, and priority is given to them based on the core task sequence, and additional tasks are shut down.

[0032] In practice, for example, a hospital's two-year-old smart food sample cabinet detected a CPU temperature of 85°C (normal threshold: 75°C), resulting in a 30% drop in work efficiency. Software data showed a 25% task queue delay. This data was identified as abnormal, and the system's data processing capacity was reduced to 70% of normal. Tasks were categorized as basic tasks (temperature monitoring, door lock control), core tasks (alarm processing), and supplementary tasks (data visualization). The core task priority sequence was as follows: Level 1 - High Temperature Alarm, Level 2 - Network Reconnection, and Level 3 - Data Retransmission. Based on the reduced processing capacity, data visualization tasks were shut down (freeing up 12% of GPU resources), and the temperature monitoring task interval was reduced from 5 minutes to 10 minutes (reducing computational load by 50%) to ensure that alarm response times remained within 200ms. When the CPU temperature dropped below 70°C, visualization tasks were gradually resumed, prioritizing loading temperature curve charts (occupying 8% of GPU resources) and pausing 3D model rendering (occupying 15% of GPU resources).

[0033] The steps for collecting task data from the smart food sample cabinet, classifying the task data into real-time tasks and deferred tasks, processing the real-time tasks in real time, and uploading the deferred tasks to the cloud are as follows: Collect task data from the smart food sample cabinet and classify the task data into real-time tasks and delayed tasks; Based on real-time tasks, the system of the intelligent food sample cabinet is called to process the real-time tasks in real time; Based on the deferred task, upload the deferred task to the cloud, call the cloud to process the deferred task, and generate a return data packet; Obtain the busy state of the system, collect the idle time of the busy state, obtain the idle time period, and send the return data packet back to the smart food sample cabinet within the idle time period.

[0034] In practice, for example, a smart food sample cabinet, used for two years at a hospital, handles both real-time tasks (temperature collection and access control) and deferred tasks (monthly report generation and log analysis). The system uses three of its four CPU cores to handle real-time tasks: Core 1 is dedicated to temperature sampling (taking 50 sensor readings every 5 minutes), Core 2 handles door lock status detection (scanning the switch signal 10 times per second), and Core 3 performs abnormal temperature alarms (response time < 1 second). Deferred tasks, such as log analysis (requiring 2 hours of CPU time), are packaged into data packets and uploaded to the cloud, where the cloud completes the analysis and generates a PDF report within 30 minutes. The system downloads reports during the idle time period between 2:00 AM and 5:00 AM, limiting bandwidth usage from 100 Mbps during the day to 10 Mbps at night. For example, one day, the system downloaded five reports (totaling 500 MB) at 3:00 AM, taking 8 minutes to complete, without affecting CPU resource allocation for real-time tasks.

[0035] An embodiment of the present invention provides a control system for an intelligent food sample cabinet, using any of the control methods for an intelligent food sample cabinet described above. The system includes the following: Scheduling prediction module 1: used to obtain historical resource learning samples and total resources of the smart food sample cabinet, obtain the resource allocation curve based on the historical resource learning samples and total resources, and generate the current prediction task scheduling mode of the smart food sample cabinet according to the resource allocation curve; Scheduling adjustment module 2: used to obtain the current usage scenario of the smart food sample cabinet, adjust the prediction task scheduling mode according to the usage scenario, and obtain the target task scheduling mode; Resource preemption module 3: Based on the target task scheduling mode, it obtains the dynamic weight of each task and monitors the dynamic weight. When the dynamic weight is higher than the preset weight threshold, the corresponding task is marked as a burst task, and the burst task automatically preempts resources. Abnormal allocation module 4: used to monitor the system of the intelligent food sample cabinet, obtain abnormal data, and dynamically adjust tasks based on the abnormal data and dynamic weights; Data processing module 5: used to collect task data of the smart food sample cabinet according to the target task scheduling mode, classify the task data into real-time tasks and delayed tasks, process the real-time tasks in real time, and upload the delayed tasks to the cloud.

[0036] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A control method for an intelligent food sample cabinet, characterized in that: The method comprises: Obtain historical resource learning samples and total resource amounts of the smart food sample cabinet, obtain a resource allocation curve based on the historical resource learning samples and the total resource amounts, and generate a prediction task scheduling mode for the current smart food sample cabinet according to the resource allocation curve; Obtaining a current usage scenario of the smart food sample cabinet, adjusting the predicted task scheduling mode according to the usage scenario, and obtaining a target task scheduling mode; Based on the target task scheduling mode, a dynamic weight of each task is obtained, and the dynamic weight is monitored. When the dynamic weight is higher than a preset weight threshold, the corresponding task is marked as a burst task, and the burst task automatically preempts resources; Monitor the system of the intelligent food sample cabinet to obtain abnormal data, and dynamically adjust the task according to the abnormal data and the dynamic weight; According to the target task scheduling mode, the task data of the smart food sample cabinet is collected, the task data is classified into real-time tasks and delayed tasks, the real-time tasks are processed in real time, and the delayed tasks are uploaded to the cloud.

2. The control method of the intelligent food sample cabinet according to claim 1 is characterized in that: The steps of obtaining historical resource learning samples of the smart food sample cabinet, obtaining a resource allocation curve based on the historical resource learning samples and the total amount of resources, and generating a prediction task scheduling mode for the current smart food sample cabinet according to the resource allocation curve are specifically as follows: Obtain historical resource learning samples of the smart food sample cabinet, construct a timeline, and expand the historical resource learning samples on the timeline to obtain a sample data change axis; According to the sample data change axis, a historical resource change law and a historical resource change value are obtained, and a historical resource change curve is generated according to the historical resource change law and the historical resource change value; Obtaining the total amount of resources, obtaining a historical resource occupancy curve based on the total amount of resources and the historical resource change curve, and generating a resource allocation curve based on the historical resource occupancy curve; The total number of tasks of the intelligent food sample retention cabinet is obtained, and a forecast task scheduling model is generated by combining the resource allocation curve and the total number of tasks.

3. The control method of the intelligent food sample cabinet according to claim 2, characterized in that: The steps of obtaining the total number of tasks of the intelligent food sample cabinet and generating a predicted task scheduling model based on the resource allocation curve and the total number of tasks are specifically as follows: Obtain the total number of tasks for the smart food sample cabinet and extract the amount of task resources required for the execution of each task; According to the resource allocation curve, the resource allocation curve is used to identify peak resources to obtain peak time periods and peak tasks corresponding to the peak time periods; Extracting the task type and resource occupancy of the peak task, and generating resource redundancy according to the task type and resource occupancy; Extracting the task importance value of each task based on the peak period to obtain idle tasks, and obtaining the idle resource amount according to the idle tasks and the task resource amount; The idle resource amount is scheduled and predicted according to the peak period, the peak task and the resource redundancy, and a predicted task scheduling mode is generated.

4. The control method of the intelligent food sample cabinet according to claim 3 is characterized in that: The steps of obtaining the current usage scenario of the smart food sample cabinet, adjusting the predicted task scheduling mode according to the usage scenario, and obtaining the target task scheduling mode are specifically as follows: Obtain the number of samples in the current smart food sample cabinet, the number of cabinet door openings and closings, the cabinet ambient temperature, and network status parameters; Obtaining usage of the smart food sample cabinet according to the number of samples in the cabinet, the number of cabinet door openings and closings, and the ambient temperature in the cabinet; According to the network status parameters, the network environment of the smart food sample cabinet is obtained, and the usage scenario of the smart food sample cabinet is obtained in combination with the usage situation and the network environment; Based on the usage scenario, the task resource requirements of the smart food sample retention cabinet are obtained, and the predicted task scheduling mode is adjusted according to the task resource requirements to obtain the target task scheduling mode.

5. The control method of the intelligent food sample storage cabinet according to claim 3 is characterized in that: Based on the target task scheduling mode, the steps of obtaining the dynamic weight of each task are specifically as follows: Based on the target task scheduling mode, the work content type of each task in the working process of the intelligent food sample cabinet is obtained; Based on the work content type, each task is evaluated for importance to obtain an importance value of each task, and an initial weight value is obtained according to the importance value; extracting the peak task type of each of the peak tasks and the peak period length of each of the peak periods based on the resource allocation curve; The initial weight values are dynamically programmed according to the peak task type and the peak period length to obtain a dynamic weight for each task.

6. The control method of the intelligent food sample storage cabinet according to claim 5, characterized in that: The dynamic weight is monitored, and when the dynamic weight is higher than a preset weight threshold, the corresponding task is marked as a burst task, and the burst task automatically preempts resources, specifically: Monitoring the dynamic weights to obtain a current weight value of each dynamic weight, and determining whether the current weight value is higher than a preset weight threshold; If it is determined that the current weight value is higher than the weight threshold, marking the current weight value as a burst weight value, and marking the task corresponding to the burst weight value as a burst task; Determine whether the current weight value is lower than or equal to a preset weight lower limit threshold; if it is determined that the current weight value is lower than the weight lower limit threshold, mark the current weight value as a buffer weight value, and mark the task corresponding to the buffer weight value as a buffer task; The buffer task is closed to release the buffer resources of the buffer task, and the buffer resources are added to the burst task.

7. The control method of the intelligent food sample storage cabinet according to claim 6, characterized in that: The steps of monitoring the system of the intelligent food sample cabinet to obtain abnormal data and dynamically regulating the task according to the abnormal data and the dynamic weight are specifically as follows: Monitor the system of the smart food sample cabinet to obtain hardware temperature data and software operation data; Based on the hardware temperature data, the current working efficiency of the system hardware is obtained, and based on the software operation data, the current fluency of the system software is obtained; Based on the current working efficiency and the current fluency, performing an abnormality assessment on the hardware temperature data and the software operation data to obtain abnormal data, and obtaining the data processing capability of the system according to the abnormal data; According to the task content type, the tasks are divided into basic tasks, core tasks and supplementary tasks; Obtaining the task priorities of the core tasks, and sorting them according to the task priorities to generate a core task sequence; According to the data processing capabilities, the basic tasks are effectively guaranteed, and according to the core task sequence, priority is given to guarantee, and the additional tasks are shut down.

8. The control method of the intelligent food sample storage cabinet according to claim 7, characterized in that: The steps of collecting task data from the smart food sample cabinet, classifying the task data into real-time tasks and deferred tasks, processing the real-time tasks in real time, and uploading the deferred tasks to the cloud are specifically as follows: Collecting task data of the intelligent food sample cabinet, and classifying the task data into real-time tasks and deferred tasks; Based on the real-time task, calling the system of the intelligent food sample cabinet to process the real-time task in real time; Based on the deferred task, uploading the deferred task to the cloud, calling the cloud to process the deferred task, and generating a return data packet; The busy state of the system is acquired, the idle time of the busy state is collected to obtain an idle time period, and the return data packet is sent back to the smart food sample cabinet within the idle time period.

9. A control system for an intelligent food sample cabinet, the system using a control method for an intelligent food sample cabinet according to any one of claims 1 to 8, characterized in that: The system comprises: Scheduling prediction module: used to obtain historical resource learning samples and total resource amount of the smart food sample cabinet, obtain a resource allocation curve based on the historical resource learning samples and the total resource amount, and generate a prediction task scheduling mode for the current smart food sample cabinet according to the resource allocation curve; Scheduling adjustment module: used to obtain the current usage scenario of the smart food sample cabinet, adjust the predicted task scheduling mode according to the usage scenario, and obtain the target task scheduling mode; Resource preemption module: used to obtain the dynamic weight of each task based on the target task scheduling mode, and monitor the dynamic weight. When the dynamic weight is higher than the preset weight threshold, the corresponding task is marked as a burst task, and the burst task automatically preempts resources; Abnormal allocation module: used to monitor the system of the intelligent food sample cabinet, obtain abnormal data, and dynamically adjust the tasks according to the abnormal data and the dynamic weight; Data processing module: used to collect task data of the smart food sample cabinet according to the target task scheduling mode, classify the task data into real-time tasks and delayed tasks, process the real-time tasks in real time, and upload the delayed tasks to the cloud.

Images