Intelligent old-age service automation supervision system

Abstract

The application relates to the technical field of smart endowment, in particular to a smart endowment service automatic supervision system, which comprises classifying the activity items of the old people into timed activities and free activities, determining the mood states and activity characteristic states of the old people in different time periods based on heart rates, establishing the relevance of the mood and the activities and marking the relevant tags, and a service recommendation module combining the real-time mood, the relevant tags and the activity tags to push the optimal services; the application can push the adaptive services by combining the activity characteristics when the old people have negative emotions, timely relieve the emotions, break the vicious circle of "emotional problems leading to physical deterioration", reduce the health risks, optimize the resource allocation according to the correlation between the mood and the service effect, avoid the waste of inefficient services, provide more targeted personalized services, and comprehensively improve the endowment service quality and the life happiness of the old people.

Description

Technical Field

[0001] This invention relates to the field of smart elderly care technology, and in particular to an automated monitoring system for smart elderly care services. Background Technology

[0002] Smart elderly care is a sensor network system and information platform for seniors living at home, in communities, and in elderly care institutions. Based on this, it provides real-time, fast, efficient, low-cost, IoT-enabled, interconnected, and intelligent elderly care services.

[0003] The prior art CN116230255A discloses a remote elderly care monitoring method, system, and electronic device, including acquiring human health data collected by a detection device, calculating elderly care monitoring data based on the human health data; acquiring health management configuration instructions based on the elderly care monitoring data, configuring corresponding home appliances based on the health management configuration instructions; and binding the detection device and home appliances to a first user of the application server's client, where the first user can be any user.

[0004] However, in the process of providing elderly care services, most service data is only analyzed in terms of physical health status, while ignoring the emotional state of the elderly. In other words, traditional service models lack the ability to analyze emotions and cannot intervene in such problems in a timely manner, which leads to the waste and misallocation of service resources. For example, when the elderly experience loss of appetite due to poor mood, traditional services may only adjust their diet from the perspective of "nutritional supplementation" without providing psychological counseling or recreational activities to address the root cause of their emotions, ultimately resulting in poor service outcomes. Summary of the Invention

[0005] The purpose of this invention is to solve the problems in the background art by proposing an automated monitoring system for smart elderly care services.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: An automated monitoring system for smart elderly care services includes: The information storage module is used to store basic information about the service population; The data monitoring module is used to collect the activity items and physiological data of the target service personnel in real time; The activity processing module is used to receive real-time activity items from target service personnel, analyze the specific performance time of each activity item, and determine the activity tag of the activity item. The activity tag includes timed activities and free activities. The state analysis module is used to acquire historical physiological data of the target service personnel, analyze the physiological data, determine the mood state of the target service personnel in each local time period, and then determine the comprehensive representative state and characteristic state of the activity item in the corresponding independent time period based on multiple local time periods. The mood state includes calm, pleasure and anger. The correlation analysis module is used to obtain the comprehensive representative status of independent time and the characteristic status of activity items, analyze the correlation between the characteristic status of activity items and the comprehensive representative status of independent time, identify the associated items, and mark the corresponding characteristic status as correlation tags. The service recommendation module is used to determine the anger state based on the current mood, then select appropriate related items based on associated tags, and finally determine the preferred service items based on activity tags.

[0007] As a further aspect of the present invention, the service population refers to the elderly population that requires elderly care supervision, and the basic information refers to the identity information and health status information of each elderly person. The identity information includes name, gender and age, and the health status information refers to the current physical health data of the elderly person.

[0008] As a further aspect of the present invention, physiological data refers to the direct quantitative representation of the elderly person's autonomic nervous system, endocrine system, and basal metabolic state, and activity items refer to the daily activities of the target service personnel. Both physiological data and activity items are measured in real time by wearable devices.

[0009] As a further aspect of the present invention, the method for determining the activity tag includes: S1: Obtain the historical information of the target service personnel, and set the unit cycle time. Divide the time in the historical information according to the unit cycle time to obtain several independent time periods. The unit cycle time is set to 1 day. Extract activities from historical information, divide activities into independent time periods, and integrate activities within an independent time period into a single set of activities. S2: Label each activity item sequentially as a target activity, identify the set of individual activities containing the target activities, extract the specific performance time of each target activity, take the average of these performance times, and label it as Ta. Then, use the formula... The time discrete value Ut is obtained, where i represents the target activity at different times, n represents the total number of target activities, and the specific time refers to the time when the target activity begins. S3: Based on the time discrete value Ut of the target activity, compare the time discrete value Ut with the baseline threshold X1. If Ut≤X1, mark the corresponding target activity as a timed activity. At the same time, obtain the time mean Ta, set the time mean Ta as the midpoint value, and determine the fixed time interval of the target activity based on the fixed duration. The fixed duration is set to 2 hours. If Ut > X1, then the corresponding target activity will be marked as a free activity. Free activities do not have a fixed time interval and can be carried out at any point in time within an independent time period.

[0010] As a further aspect of the present invention, the method for determining the comprehensive representative state includes: Acquire historical physiological data of the target service personnel, divide the physiological data according to independent time based on the collection time of the physiological data, and integrate the physiological data within the same independent time period into a data set; Choose any independent time, mark this independent time as the specified analysis time, obtain the data set corresponding to the specified analysis time, set the unit time, and divide the specified analysis time into multiple local time periods according to the unit time; Calculate the independent mood coefficient in a local time period, and compare the independent mood coefficient Xq with the mood thresholds Y1 and Y2. If Xq < Y1, the mood state of the corresponding local time period is marked as calm; if Y1 ≤ Xq < Y2, the mood state of the corresponding local time period is marked as happy; if Xq ≥ Y2, the mood state of the corresponding local time period is marked as angry, where Y1 > Y2. The above method is used to process all local data segments within an independent time period to determine the mood state of each local time period; Then, within an independent time period, the frequency of each mood state is counted, and the mood state with the most occurrences is marked as the comprehensive representative state for the corresponding independent time period.

[0011] As a further aspect of the present invention, the method for calculating the independent mood coefficient includes: Identify the collection time of physiological data in the dataset, and mark continuous physiological data within the same local time period as local data segments according to the time range of local time periods. Then, set time as the horizontal axis and physiological data as the vertical axis, set up a plane coordinate system, and mark the local data segments on the plane coordinate system and connect them linearly to obtain the data change curve. Obtain baseline data for physiological data and mark the baseline data on a plane coordinate system to generate a baseline line. The baseline data refers to the average value of the target service personnel in a calm state. In the plane coordinate system, identify all peaks and troughs in the data change curve. First, count the number of peaks and troughs and mark them as the number of peaks and troughs DL. Subtract the baseline data from each peak and trough value to obtain the fluctuation value. The fluctuation value is taken as a positive number. Then, calculate the mean of the fluctuation value and mark the mean result as the state characteristic value Zt. Using formula The independent mood coefficient Xq for this local time period is obtained, where a is the base and 0 < a < 1. These are the weighting coefficients.

[0012] As a further aspect of the present invention, when statistically analyzing mood states over independent time periods, it is necessary to identify the normal sleep time of the target service personnel and exclude the local time periods corresponding to the normal sleep time from the statistical analysis of mood states over the corresponding independent time periods. That is, when statistically analyzing mood states, the local time periods corresponding to the normal sleep time are not included in the statistical process of mood states. Here, normal sleep time refers to the nighttime sleep time of the target service personnel.

[0013] As a further aspect of the present invention, the method for determining the characteristic state of an activity item includes: Obtain the set of individual activities corresponding to independent time periods. Based on the specific performance time of each activity item in the set of individual activities, and taking the time length of the local time period as the benchmark, divide the specific performance time into several activity periods according to the unit time, starting from the end time of the specific performance time. Specifically, if an activity period is shorter than a unit time, the corresponding activity period will be deleted. Then, the physiological data corresponding to each activity period is obtained, and the physiological data is processed according to the above method to obtain the independent mood coefficient of each activity period. In an independent time, the independent mood coefficients of the same activity item in multiple activity periods are averaged, and the result of the averaging is marked as the mood feature value. Then, the mood feature value is compared with the mood thresholds Y1 and Y2 respectively to determine the characteristic state of this activity item.

[0014] As a further aspect of the present invention, the method for determining associated items and associated tags includes: Arbitrarily select an activity item m, identify the independent time in which activity item m is located, count the number of this independent time and mark it as the occurrence count Rc, obtain the comprehensive representative state corresponding to the independent time, if the comprehensive representative state of the independent time is consistent with the characteristic state of activity item m, then generate a signal with the same frequency; otherwise, if the comprehensive representative state of the independent time is inconsistent with the characteristic state of activity item m, then generate an irrelevant signal. The number of times the same frequency signal of activity item m appears is counted and marked as the same frequency quantity Rt. The same frequency ratio F of activity item is obtained by using the formula F=Rt÷Rc. The same frequency ratio F is compared with the ratio coefficient X2. If F<X2, activity item m is marked as an irrelevant item. Otherwise, if F≥X2, activity item m is marked as an associated item, and the corresponding feature state is marked as an associated label.

[0015] As a further aspect of the present invention, the preferred method for determining service items includes: Acquire physiological data at the current local time and calculate the independent mood coefficient at this time. The mood state at the current local time is determined by the independent mood coefficient. Then, obtain the mood status of all local time periods within the day. If anger is present in the mood status, identify the related items in all activities and take the related items with the tag of happiness. Based on the current specific time point and the activity tag of the related items, select the time-appropriate related items and take these related items as the current preferred service items to recommend items to the target service personnel. Among them, time-adapted associated projects refer to the following: if the activity tag of the associated project is free activity, the corresponding associated project will be directly marked as a preferred service project; if the activity tag of the associated project is timed activity, the corresponding fixed time interval will be obtained; if the fixed time interval contains the current specific time point, the corresponding associated project will be marked as a preferred service project; if the fixed time interval does not contain the current specific time point, the corresponding associated project will not be marked as a preferred service project.

[0016] Compared with existing technologies, the advantages of this invention are: This invention provides accurate and timely data support for supervision by centrally managing information on the service population and monitoring the activities and physiological dynamics of the elderly in real time, thus achieving dynamic and precise supervision and avoiding risk omissions due to missing or delayed information. At the same time, it breaks away from the one-sidedness of traditional elderly care that only focuses on physical health data and ignores emotional state. It can accurately judge the mood of the elderly at different times through physiological data, establish the correlation between mood and activity, fill the gap in emotional supervision, and build a "physical and mental dual-dimensional" supervision system. This invention also addresses the issue by providing tailored services based on the elderly’s activity characteristics when they experience negative emotions, thus timely alleviating their emotions, breaking the vicious cycle of “emotional problems leading to physiological deterioration,” reducing health risks, optimizing resource allocation based on the correlation between mood and service effectiveness, avoiding inefficient service waste, providing more targeted and personalized services, and comprehensively improving the quality of elderly care services and the well-being of the elderly. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0019] Reference Figure 1 An automated monitoring system for smart elderly care services includes an information storage module, a data monitoring module, an activity processing module, a status analysis module, a correlation analysis module, and a service recommendation module. The information storage module is used to store the basic information of the service population. In this embodiment, the service population refers to the elderly population that needs to be supervised. The basic information refers to the identity information and health status information of each elderly person. Furthermore, the identity information includes information such as name, gender and age, and the health status information refers to the current physical health data of the elderly person. Then, the information storage module establishes a one-way communication connection with the status analysis module and the activity processing module respectively. The data monitoring module is used to collect the activity items and physiological data of the target service personnel in real time, and transmit the collected activity items and physiological data to the information storage module for real-time data processing. At the same time, the data monitoring module also establishes a one-way communication connection with the service recommendation module. The target service personnel refer to the elderly people who need to be monitored for elderly care services. It should be further explained that physiological data refers to the direct quantitative data reflecting the autonomic nervous system, endocrine system and basal metabolic state of the elderly, including data such as heart rate and blood pressure. In this embodiment, physiological data refers to heart rate. Activity items refer to the daily activities of the target service personnel, including social activities, exercise and intellectual activities (such as playing chess). Both physiological data and activity items are measured in real time by wearable devices, including smart bracelets and smartwatches. The activity processing module receives real-time activity items from target service personnel and analyzes their historical data. By combining this data with the current activity items, the module determines the nature of each activity. Specifically, the methods for determining the nature of each activity include: S1: Obtain the historical information of the target service personnel, and set the unit period time. Divide the time in the historical information according to the unit period time to obtain several independent time periods. The specific duration of the unit cycle time is set by those skilled in the art based on big data experience. In this embodiment, the unit cycle time is set to 1 day. Extract activities from historical information, divide activities into independent time periods, and integrate activities within an independent time period into a single set of activities. S2: Randomly select an activity item. Taking this activity item as an example, mark it as the target activity. Identify the set of individual activities to which the target activity belongs, and extract the specific performance time of the target activity. Take the average of the specific performance times and mark it as Ta. Then use the formula... The time discrete value Ut is obtained, where i represents the target activity at different times, n represents the total number of target activities, and the specific time refers to the time when the target activity starts, such as 9:30. S3: Based on the time discrete value Ut of the target activity, compare the time discrete value Ut with the benchmark threshold X1. If Ut≤X1, mark the corresponding target activity as a timed activity. At the same time, obtain the time mean Ta, set the time mean Ta as the midpoint value, and determine the fixed time interval of the target activity based on the fixed duration. Furthermore, the specific value of the fixed duration is set by those skilled in the art based on big data experience. In this embodiment, the fixed duration is set to 2 hours. If Ut > X1, then the activity tag of the corresponding target activity will be marked as a free activity. Free activities do not have a fixed time interval and can be carried out at any point in time within an independent time period. Furthermore, the specific value of the benchmark threshold X1 is obtained by those skilled in the art based on big data calculations; Mark all activities as target activities in sequence, and determine the activity nature of each activity, where activity nature refers to free activities and timed activities; Then, a one-way communication connection is established between the activity processing module and the service recommendation module, and the activity nature of the activity item is transmitted to the service recommendation module; The state analysis module is used to acquire historical physiological data of the target service personnel and analyze the physiological data to determine the target service personnel's mood state. Mood state includes calm, pleasure, and anger. Specific methods for determining mood state include: SS1: Acquire historical physiological data of the target service personnel, divide the physiological data according to independent time based on the collection time of the physiological data, and integrate the physiological data within the same independent time period into a data set; Choose any independent time, mark this independent time as the specified analysis time, obtain the data set corresponding to the specified analysis time, set the unit time, and divide the specified analysis time into multiple local time periods according to the unit time. The specific length of the unit time is set by those skilled in the art based on big data experience. In this embodiment, the unit time is set to 1 hour. SS2: Identify the collection time of physiological data in the dataset, and mark continuous physiological data within the same local time period as local data segments according to the time range of local time periods. Then, set time as the horizontal axis and physiological data as the vertical axis, set up a plane coordinate system, and mark the local data segments on the plane coordinate system and connect them linearly to obtain the data change curve. SS3: Obtain baseline data for physiological data and mark the baseline data on a plane coordinate system to generate a baseline line. The baseline data refers to the average value of the target service personnel in a calm state. The specific value of the baseline data is obtained by those skilled in the art based on big data calculations. In the plane coordinate system, identify all peaks and troughs in the data change curve. First, count the number of peaks and troughs and mark them as the number of peaks and troughs DL. Subtract the baseline data from each peak and trough value to obtain the fluctuation value. The fluctuation value is taken as a positive number. Then, calculate the mean of the fluctuation value and mark the mean result as the state characteristic value Zt. Then use the formula The independent mood coefficient Xq for this local time period is obtained, where a is the base and 0 < a < 1. For the weighting coefficients, further, a and The specific values ​​are set by those skilled in the art based on their experience with big data; SS4: Compare the independent mood coefficient Xq with mood thresholds Y1 and Y2. If Xq < Y1, the mood state of the corresponding local time period is marked as calm. If Y1 ≤ Xq < Y2, the mood state of the corresponding local time period is marked as happy. If Xq ≥ Y2, the mood state of the corresponding local time period is marked as angry. Furthermore, the specific values ​​of mood thresholds Y1 and Y2 are obtained by those skilled in the art based on big data calculations, and Y1 > Y2. The above method is used to process all local data segments within an independent time period to determine the mood state of each local time period; Then, within an independent time period, the frequency of each mood state is counted, and the mood state with the most occurrences is marked as the comprehensive representative state for the corresponding independent time period; It should be further explained that when statistically analyzing mood status during independent time periods, it is necessary to identify the normal sleep time of the target service personnel and exclude the local time periods corresponding to the normal sleep time from the statistics of mood status during the corresponding independent time periods. That is, when statistically analyzing mood status, the local time periods corresponding to the normal sleep time are not included in the statistical process of mood status. Here, normal sleep time refers to the nighttime sleep time of the target service personnel. SS5: Obtain the set of individual activities corresponding to this independent time, and based on the specific performance time of each activity item in the set of individual activities, divide the specific performance time into several activity periods according to the length of the local time period, that is, according to the unit time, starting from the end time of the specific performance time. It should be further noted that if an activity period is shorter than the unit time, the corresponding activity period will be deleted. Then, the physiological data corresponding to each activity period is obtained, and the physiological data is processed according to the above method to obtain the independent mood coefficient of each activity period. In an independent time, the independent mood coefficients of the same activity item in multiple activity periods are averaged, and the result of the averaging is marked as the mood feature value. Then, the mood feature value is compared with the mood thresholds Y1 and Y2 respectively to determine the characteristic state of this activity item. Then, a one-way communication connection is established between the state analysis module and the correlation analysis module; The correlation analysis module is used to obtain the comprehensive representative status of independent time periods and the characteristic status of activity items, and to analyze the correlation between the characteristic status of activity items and the comprehensive representative status of independent time periods. Specific correlation analysis methods include: Arbitrarily select an activity item m, identify the independent time in which activity item m is located, count the number of this independent time and mark it as the occurrence count Rc, obtain the comprehensive representative state corresponding to the independent time, if the comprehensive representative state of the independent time is consistent with the characteristic state of activity item m, then generate a signal with the same frequency; otherwise, if the comprehensive representative state of the independent time is inconsistent with the characteristic state of activity item m, then generate an irrelevant signal. The number of times the same frequency signal of activity item m appears is counted and marked as the same frequency quantity Rt. The same frequency ratio F of activity item is obtained by using the formula F=Rt÷Rc. The same frequency ratio F is compared with the ratio coefficient X2. If F<X2, activity item m is marked as an irrelevant item. Otherwise, if F≥X2, activity item m is marked as an associated item, and the corresponding feature state is marked as an associated label. The specific value of the ratio coefficient X2 is obtained by those skilled in the art based on big data calculation. Then, a one-way communication connection is established between the association analysis module and the service recommendation module; The service recommendation module is used to acquire physiological data at the current local time and calculate the independent mood coefficient at that time. The independent mood coefficient determines the mood state at the current local time. Then, it acquires the mood state at all local times within the day. If anger is present in the mood state, it identifies related items among all activities and takes the related items with the label of "pleasure". Based on the current specific time point and the activity label of the related items, it selects time-adapted related items and takes these related items as the current preferred service items. It then recommends these items to the target service personnel. Time-adapted related items mean: if the activity label of the related item is "free activity", the corresponding related item is directly marked as a preferred service item; if the activity label of the related item is "scheduled activity", the corresponding fixed time interval is obtained. If there is a fixed time interval that includes the current specific time point, the corresponding related item is marked as a preferred service item; if the fixed time interval does not include the current specific time point, the corresponding related item is not marked as a preferred service item.

[0020] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An automated monitoring system for smart elderly care services, characterized in that, include: The information storage module is used to store basic information about the service population; The data monitoring module is used to collect the activity items and physiological data of the target service personnel in real time; The activity processing module is used to receive real-time activity items from target service personnel, analyze the specific performance time of each activity item, and determine the activity tag of the activity item. The activity tag includes timed activities and free activities. The state analysis module is used to acquire historical physiological data of the target service personnel, analyze the physiological data, determine the mood state of the target service personnel in each local time period, and then determine the comprehensive representative state and characteristic state of the activity item in the corresponding independent time period based on multiple local time periods. The mood state includes calm, pleasure and anger. The correlation analysis module is used to obtain the comprehensive representative status of independent time and the characteristic status of activity items, analyze the correlation between the characteristic status of activity items and the comprehensive representative status of independent time, identify the associated items, and mark the corresponding characteristic status as correlation tags. The service recommendation module is used to determine the anger state based on the current mood, then select appropriate related items based on associated tags, and finally determine the preferred service items based on activity tags.

2. The automated monitoring system for smart elderly care services according to claim 1, characterized in that, The service population refers to the elderly population that requires elderly care supervision. The basic information refers to the identity information and health status information of each elderly person. The identity information includes name, gender and age, and the health status information refers to the current physical health data of the elderly person.

3. The automated monitoring system for smart elderly care services according to claim 1, characterized in that, Physiological data refers to the direct quantitative representation of the elderly person's autonomic nervous system, endocrine system, and basal metabolic state. Activity items refer to the daily activities of the target service personnel. Both physiological data and activity items are measured in real time by wearable devices.

4. The automated monitoring system for smart elderly care services according to claim 1, characterized in that, Methods for determining activity tags include: S1: Obtain the historical information of the target service personnel, and set the unit cycle time. Divide the time in the historical information according to the unit cycle time to obtain several independent time periods. The unit cycle time is set to 1 day. Extract activities from historical information, divide activities into independent time periods, and integrate activities within an independent time period into a single set of activities; S2: Label each activity item sequentially as a target activity, identify the set of individual activities containing the target activities, extract the specific performance time of each target activity, take the average of these performance times, and label it as Ta. Then, use the formula... The time discrete value Ut is obtained, where i represents the target activity at different times, n represents the total number of target activities, and the specific time refers to the time when the target activity begins. S3: Based on the time discrete value Ut of the target activity, compare the time discrete value Ut with the baseline threshold X1. If Ut≤X1, mark the corresponding target activity as a timed activity. At the same time, obtain the time mean Ta, set the time mean Ta as the midpoint value, and determine the fixed time interval of the target activity based on the fixed duration. The fixed duration is set to 2 hours. If Ut > X1, then the corresponding target activity will be marked as a free activity. Free activities do not have a fixed time interval and can be carried out at any point in time within an independent time period.

5. The automated monitoring system for smart elderly care services according to claim 1, characterized in that, The methods for determining the comprehensive representative state include: Acquire historical physiological data of the target service personnel, divide the physiological data according to independent time based on the collection time of the physiological data, and integrate the physiological data within the same independent time period into a data set; Choose any independent time, mark this independent time as the specified analysis time, obtain the data set corresponding to the specified analysis time, set the unit time, and divide the specified analysis time into multiple local time periods according to the unit time; Calculate the independent mood coefficient in a local time period, and compare the independent mood coefficient Xq with the mood thresholds Y1 and Y2. If Xq < Y1, the mood state of the corresponding local time period is marked as calm; if Y1 ≤ Xq < Y2, the mood state of the corresponding local time period is marked as happy; if Xq ≥ Y2, the mood state of the corresponding local time period is marked as angry, where Y1 > Y2. The above method is used to process all local data segments within an independent time period to determine the mood state of each local time period; Then, within an independent time period, the frequency of each mood state is counted, and the mood state with the most occurrences is marked as the comprehensive representative state for the corresponding independent time period.

6. The automated monitoring system for smart elderly care services according to claim 5, characterized in that, The methods for calculating the independent mood coefficient include: Identify the collection time of physiological data in the dataset, and mark continuous physiological data within the same local time period as local data segments according to the time range of local time periods. Then, set time as the horizontal axis and physiological data as the vertical axis, set up a plane coordinate system, and mark the local data segments on the plane coordinate system and connect them linearly to obtain the data change curve. Obtain baseline data for physiological data and mark the baseline data on a plane coordinate system to generate a baseline line. The baseline data refers to the average value of the target service personnel in a calm state. In the plane coordinate system, identify all peaks and troughs in the data change curve. First, count the number of peaks and troughs and mark them as the number of peaks and troughs DL. Subtract the baseline data from each peak and trough value to obtain the fluctuation value. The fluctuation value is taken as a positive number. Then, calculate the mean of the fluctuation value and mark the mean result as the state characteristic value Zt. Using formula The independent mood coefficient Xq for this local time period is obtained, where a is the base and 0 < a < 1. These are the weighting coefficients.

7. The automated monitoring system for smart elderly care services according to claim 5, characterized in that, When statistically analyzing mood states over independent periods of time, it is necessary to identify the normal sleep time of the target service personnel and exclude the local time periods corresponding to the normal sleep time from the statistical analysis of mood states within the corresponding independent time periods. In other words, when statistically analyzing mood states, the local time periods corresponding to the normal sleep time are excluded from the statistical process of mood states. Here, normal sleep time refers to the nighttime sleep time of the target service personnel.

8. The automated monitoring system for smart elderly care services according to claim 7, characterized in that, Methods for determining the characteristic states of an activity item include: Obtain the set of individual activities corresponding to independent time periods. Based on the specific performance time of each activity item in the set of individual activities, and taking the time length of the local time period as the benchmark, divide the specific performance time into several activity periods according to the unit time, starting from the end time of the specific performance time. Specifically, if an activity period is shorter than a unit time, the corresponding activity period will be deleted. Then, the physiological data corresponding to each activity period is obtained, and the physiological data is processed according to the above method to obtain the independent mood coefficient of each activity period. In an independent time, the independent mood coefficients of the same activity item in multiple activity periods are averaged, and the result of the averaging is marked as the mood feature value. Then, the mood feature value is compared with the mood thresholds Y1 and Y2 respectively to determine the characteristic state of this activity item.

9. The automated monitoring system for smart elderly care services according to claim 1, characterized in that, The methods for determining associated items and associated tags include: Arbitrarily select an activity item m, identify the independent time in which activity item m is located, count the number of this independent time and mark it as the occurrence count Rc, obtain the comprehensive representative state corresponding to the independent time, if the comprehensive representative state of the independent time is consistent with the characteristic state of activity item m, then generate a signal with the same frequency; otherwise, if the comprehensive representative state of the independent time is inconsistent with the characteristic state of activity item m, then generate an irrelevant signal. The number of times the same frequency signal of activity item m appears is counted and marked as the same frequency quantity Rt. The same frequency ratio F of activity item is obtained by using the formula F=Rt÷Rc. The same frequency ratio F is compared with the ratio coefficient X2. If F<X2, activity item m is marked as an irrelevant item. Otherwise, if F≥X2, activity item m is marked as an associated item, and the corresponding feature state is marked as an associated label.

10. The automated monitoring system for smart elderly care services according to claim 1, characterized in that, The methods for determining preferred service items include: Acquire physiological data at the current local time and calculate the independent mood coefficient at this time. The mood state at the current local time is determined by the independent mood coefficient. Then, obtain the mood status of all local time periods within the day. If anger is present in the mood status, identify the related items in all activities and take the related items with the tag of happiness. Based on the current specific time point and the activity tag of the related items, select the time-appropriate related items and take these related items as the current preferred service items to recommend items to the target service personnel. Among them, time-adapted associated projects refer to the following: if the activity tag of the associated project is free activity, the corresponding associated project will be directly marked as a preferred service project; if the activity tag of the associated project is timed activity, the corresponding fixed time interval will be obtained; if the fixed time interval contains the current specific time point, the corresponding associated project will be marked as a preferred service project; if the fixed time interval does not contain the current specific time point, the corresponding associated project will not be marked as a preferred service project.

Images