CN114973596A - Individual-living old people monitoring system and method based on federal multi-source data analysis - Google Patents

Abstract

The invention discloses a monitoring system and a monitoring method for the elderly living alone based on federal multi-source data analysis, wherein a controller is used for carrying out power consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm analysis; when any one of the analysis data has an alarm condition, the controller controls the ammeter, the water flow sensor, the camera, the carbon monoxide sensor, the temperature and humidity sensor and the door lock sensor to improve the monitoring sensitivity; the controller performs power utilization data, water utilization data, indoor video detection and indoor danger alarm analysis again according to data collected by the ammeter, the water flow sensor, the camera, the carbon monoxide sensor, the temperature and humidity sensor and the door lock sensor after monitoring sensitivity is improved; and the controller predicts the alarm level according to the re-analysis result and starts an alarm scheme according to the alarm level prediction result. The method uses the electric energy and water consumption data from families to judge whether the indoor has abnormal behaviors or not through various data sources.

Description

A monitoring system and method for the elderly living alone based on federal multi-source data analysis

technical field

The invention relates to the technical field of data collection and processing, in particular to a single-living elderly monitoring system and method based on federal multi-source data analysis.

Background technique

The statements in this section merely provide background related to the present disclosure and do not necessarily constitute prior art.

As the age of the elderly increases, their mobility and thinking ability will decline to varying degrees, and they are prone to accidents such as falls and sudden illnesses. Especially for the deaf-mute elderly, they are more likely to face some unexpected situations than the general elderly, and we need to pay extra attention. When children have full-time jobs, they don't have the time and energy to spend on caring for the elderly. At present, the cost of hiring a full-time carer or nanny is not small, and it is not affordable for every family. Even for financially capable families, as the population ages further, care workers will become in short supply.

At this stage, most of the nursing equipment for the elderly is used to detect the posture of the elderly. The data source of this method is relatively single, and in some cases, it is easy to make mistakes in judgment, and there is a risk of invasion of the privacy of the elderly.

SUMMARY OF THE INVENTION

In order to solve the deficiencies of the prior art, the present invention provides a single-living elderly monitoring system and method based on federated multi-source data analysis; the present invention combines the idea of federated learning, fixes the data at the detection end for local analysis, and only analyzes the judgment results. Judgment is made through network aggregation. At the same time, the present invention uses electric energy data, water consumption data, etc. from households, and judges whether there is abnormal behavior in the room through a variety of data sources.

In a first aspect, the present invention provides a monitoring system for the elderly living alone based on federal multi-source data analysis;

A monitoring system for the elderly living alone based on federal multi-source data analysis, including: a controller;

The controller, according to the data collected by the electricity meter, the water flow sensor, the camera, the carbon monoxide sensor, the temperature and humidity sensor and the door lock sensor, performs electricity consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm analysis;

When any alarm situation occurs in the analysis data, the controller controls the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor and door lock sensor to improve the monitoring sensitivity;

According to the data collected by the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor, and door lock sensor after the monitoring sensitivity is improved, the controller conducts electricity consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm again. analyze;

The controller predicts the alarm level according to the result of the re-analysis, and starts the alarm scheme according to the prediction result of the alarm level.

In a second aspect, the present invention provides a monitoring method for the elderly living alone based on federal multi-source data analysis;

A method of monitoring older adults living alone based on federal multi-source data analysis, including:

The controller, according to the data collected by the electricity meter, the water flow sensor, the camera, the carbon monoxide sensor, the temperature and humidity sensor and the door lock sensor, analyzes the electricity consumption data, the water consumption data analysis, the indoor video detection analysis and the indoor danger alarm analysis;

When any alarm situation occurs in the analysis data, the controller controls the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor and door lock sensor to improve the monitoring sensitivity;

According to the data collected by the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor, and door lock sensor after the monitoring sensitivity is improved, the controller conducts electricity consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm again. analyze;

The controller predicts the alarm level according to the result of the re-analysis, and starts the alarm scheme according to the prediction result of the alarm level.

Compared with the prior art, the beneficial effects of the present invention are:

Compared with the previous single data source method, this system conducts analysis through multiple data sources. The data analysis part of the present invention is all local, and there is no need to upload the original data through the Internet, only the analysis result needs to be uploaded.

The invention has rich data sources, can realize the prediction and monitoring of various family risks, and can effectively improve the accuracy of risk discovery by joint analysis. The present invention does not need to upload the original data through the network, and by changing the analysis mode, the privacy information of the user is protected as much as possible, and the use is more secure.

Description of drawings

The accompanying drawings forming a part of the present invention are used to provide further understanding of the present invention, and the exemplary embodiments of the present invention and their descriptions are used to explain the present invention, and do not constitute an improper limitation of the present invention.

1 is a hardware connection frame diagram of Embodiment 1;

FIG. 2 is a schematic diagram of the connection of the internal analysis, detection and early warning model of the controller according to the first embodiment.

Detailed ways

It should be noted that the following detailed description is exemplary and intended to provide further explanation of the invention. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

It should be noted that the terminology used herein is for the purpose of describing specific embodiments only, and is not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly dictates otherwise, the singular is intended to include the plural as well, furthermore, it is to be understood that the terms "including" and "having" and any conjugations thereof are intended to cover the non-exclusive A process, method, system, product or device comprising, for example, a series of steps or units is not necessarily limited to those steps or units expressly listed, but may include those steps or units not expressly listed or for such processes, methods, Other steps or units inherent to the product or equipment.

Embodiments of the invention and features of the embodiments may be combined with each other without conflict.

All data acquisition in this embodiment is based on compliance with laws and regulations and the user's consent, and the legal application of the data.

Example 1

This embodiment provides a monitoring system for the elderly living alone based on federal multi-source data analysis;

A monitoring system for the elderly living alone based on federal multi-source data analysis, including: a controller;

The controller, according to the data collected by the electricity meter, the water flow sensor, the camera, the carbon monoxide sensor, the temperature and humidity sensor and the door lock sensor, performs electricity consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm analysis;

When any alarm situation occurs in the analysis data, the controller controls the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor and door lock sensor to improve the monitoring sensitivity;

According to the data collected by the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor, and door lock sensor after the monitoring sensitivity is improved, the controller conducts electricity consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm again. analyze;

The controller predicts the alarm level according to the result of the re-analysis, and starts the alarm scheme according to the prediction result of the alarm level.

Further, as shown in FIG. 1 , the controller is respectively connected with an electric meter, a water flow sensor, a camera, a carbon monoxide sensor, a temperature and humidity sensor and a door lock sensor.

Further, the electricity meter is used to collect electricity consumption data of the elderly living alone;

The water flow sensor is used to collect water consumption data of the elderly living alone;

The camera is used to collect indoor video data of the elderly living alone;

The carbon monoxide sensor is used to collect the carbon monoxide concentration in the kitchen of the elderly living alone;

The temperature and humidity sensor is used to collect temperature data and humidity data of the bedroom of the elderly living alone;

The door lock sensor is used for collecting open and closed state data of the door for the elderly living alone.

Further, conduct electricity consumption data analysis; specifically include:

The trained power consumption data analysis model is used to analyze the real-time power consumption data collected by the meter, and output the power consumption data analysis results.

Further, the power consumption data analysis model after the training; the training step includes:

Build the first convolutional neural network;

constructing a first training set; the first training set is the historical electricity consumption data of all elderly people in the whole community with known normal or abnormal labels;

The first convolutional neural network is trained by using the first training set, and the trained first convolutional neural network is obtained;

constructing a second training set; the second training set is the historical electricity consumption data of the elderly to be monitored with known normal or abnormal labels;

Using the second training set, the trained first convolutional neural network is trained to obtain the retrained first convolutional neural network; the retrained first convolutional neural network is used as the power consumption data after training Analytical model.

Further, carry out water use data analysis; specifically include:

The trained water consumption data analysis model is used to analyze the real-time water consumption data collected by the water flow sensor, and output the water consumption data analysis results.

Further, the water data analysis model after the training; the training step includes:

Build a second convolutional neural network;

constructing a third training set; the third training set is the historical water consumption data of all the elderly in the whole community with known normal or abnormal labels;

The second convolutional neural network is trained by using the third training set, and the trained second convolutional neural network is obtained;

constructing a fourth training set; the fourth training set is the historical water consumption data of the elderly to be monitored with known normal or abnormal labels;

The fourth training set is used to train the second convolutional neural network after training to obtain the second convolutional neural network after retraining; the second convolutional neural network after retraining is used as the data analysis of water consumption after training Model.

Further, the indoor video detection and analysis; specifically include:

The indoor video detection model after training is used to detect and analyze the indoor video data collected by the cameras installed indoors, and the indoor video detection and analysis results are obtained.

Further, the indoor video detection model after the training; the specific training process includes:

Build a third convolutional neural network;

constructing a fifth training set; the fifth training set is the historical indoor activity videos of all elderly people in the whole community with known normal or abnormal labels;

The third convolutional neural network is trained by using the fifth training set, and the trained third convolutional neural network is obtained;

constructing a sixth training set; the sixth training set is the historical indoor activity videos of the elderly to be monitored with known normal or abnormal labels;

The sixth training set is used to train the third convolutional neural network after training, and the third convolutional neural network after retraining is obtained; the third convolutional neural network after retraining is used as the indoor video detection after training. Model.

Further, the indoor danger alarm analysis; specifically includes:

Using the trained indoor danger alarm model, the kitchen carbon monoxide concentration collected by the carbon monoxide sensor, the bedroom temperature and humidity collected by the temperature and humidity sensor, and the door lock opening and closing states corresponding to different time points collected by the door lock sensor are detected and analyzed, and the indoor danger is obtained. Alarm analysis results.

Further, the indoor danger alarm model after the training; the training process includes:

Build the fourth convolutional neural network;

Constructing a seventh training set; the seventh training set is the carbon monoxide concentration in the kitchen, the temperature and humidity of the bedroom, and the open and closed states of the door locks corresponding to different time points of all the elderly in the whole community with known normal or abnormal labels;

The seventh training set is used to train the fourth convolutional neural network, and the trained fourth convolutional neural network is obtained;

Constructing an eighth training set; the eighth training set is the kitchen carbon monoxide concentration, bedroom temperature and humidity, and door lock open and closed states corresponding to different time points of the elderly to be monitored with known normal or abnormal labels;

The eighth training set is used to train the fourth convolutional neural network after training, and the fourth convolutional neural network after retraining is obtained; the fourth convolutional neural network after retraining is used as the indoor danger alarm after training Model.

Further, when any alarm situation occurs in the analysis data, the controller controls the electric meter, the water flow sensor, the camera, the carbon monoxide sensor, the temperature and humidity sensor and the door lock sensor to improve the monitoring sensitivity; Boost according to the preset boost.

Further, as shown in FIG. 2 , the controller predicts the alarm level according to the result of the reanalysis, and starts the alarm scheme according to the prediction result of the alarm level; specifically, it includes:

The power consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm analysis results are all input into the trained early warning level classification model, and the early warning level is output.

Further, the trained warning level classification model; the training process includes:

Build a support vector machine classifier;

Constructing a ninth training set; the ninth training set is the analysis results of electricity consumption data, water data analysis, indoor video detection analysis and indoor danger alarm analysis of known warning level classification labels;

The ninth training set is input into the support vector machine classifier, the classifier is trained, the trained support vector machine classifier is obtained, and the trained support vector machine classifier is used as the trained early warning level classification model.

Further, according to the prediction result of the alarm level, start the alarm scheme; including:

If it is "no risk" level, no alarm;

If it is at the level of "potential risk", send a text message reminder to the mobile terminal of the family members of the elderly living alone;

If it is "at risk", call the emergency contact directly.

The present invention mainly uses user electric box data, user water data, indoor video data, indoor gas data, indoor temperature and humidity data, and door and window lock safety detection data.

All the data analysis and data collection of the present invention are integrated into one device, and when the analysis is performed, the analysis is performed locally on the device, and the data does not need to be uploaded through the network.

When the system is initialized, each analysis model uses the local machine learning model and user data for training and analysis. All the machine learning models used are pre-trained with common data, and after being deployed to the user's home, they are retrained so that the machine learning model can adapt to the specificity of the user's home habits.

After the emergency contact is confirmed, emergency calls such as 110 or 120 can be dialed through the emergency contact notification module according to the plan and emergency level.

Embodiment 2

This embodiment provides a monitoring method for the elderly living alone based on federal multi-source data analysis;

A method of monitoring older adults living alone based on federal multi-source data analysis, including:

The controller, according to the data collected by the electricity meter, the water flow sensor, the camera, the carbon monoxide sensor, the temperature and humidity sensor and the door lock sensor, analyzes the electricity consumption data, the water consumption data analysis, the indoor video detection analysis and the indoor danger alarm analysis;

When any alarm situation occurs in the analysis data, the controller controls the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor and door lock sensor to improve the monitoring sensitivity;

According to the data collected by the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor and door lock sensor after the monitoring sensitivity is improved, the controller performs electricity data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm again. analyze;

The controller predicts the alarm level according to the result of the re-analysis, and starts the alarm scheme according to the prediction result of the alarm level.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims (10)

1. A monitoring system for the elderly living alone based on federal multi-source data analysis, characterized in that it includes: a controller; The controller, according to the data collected by the electricity meter, the water flow sensor, the camera, the carbon monoxide sensor, the temperature and humidity sensor and the door lock sensor, performs electricity consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm analysis; When any alarm situation occurs in the analysis data, the controller controls the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor and door lock sensor to improve the monitoring sensitivity; According to the data collected by the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor, and door lock sensor after the monitoring sensitivity is improved, the controller conducts electricity consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm again. analyze; The controller predicts the alarm level according to the result of the re-analysis, and starts the alarm scheme according to the prediction result of the alarm level. 2. The monitoring system for the elderly living alone based on federal multi-source data analysis according to claim 1, wherein the controller is respectively connected with an electric meter, a water flow sensor, a camera, a carbon monoxide sensor, a temperature and humidity sensor, and a door lock sensor. device connection. 3. The monitoring system for the elderly living alone based on federal multi-source data analysis according to claim 2, wherein the electricity meter is used to collect the electricity consumption data of the elderly living alone; The water flow sensor is used to collect water consumption data of the elderly living alone; The camera is used to collect indoor video data of the elderly living alone; The carbon monoxide sensor is used to collect the carbon monoxide concentration in the kitchen of the elderly living alone; The temperature and humidity sensor is used to collect temperature data and humidity data of the bedroom of the elderly living alone; The door lock sensor is used for collecting open and closed state data of the door for the elderly living alone. 4. The monitoring system for the elderly living alone based on federal multi-source data analysis as claimed in claim 1, wherein the analysis of electricity consumption data is performed; specifically, it comprises: Using the power consumption data analysis model after training, analyze the real-time power consumption data collected by the meter, and output the power consumption data analysis results; The power consumption data analysis model after the training; the training step includes: Build the first convolutional neural network; constructing a first training set; the first training set is the historical electricity consumption data of all elderly people in the whole community with known normal or abnormal labels; The first convolutional neural network is trained by using the first training set, and the trained first convolutional neural network is obtained; constructing a second training set; the second training set is the historical electricity consumption data of the elderly to be monitored with known normal or abnormal labels; Using the second training set, the trained first convolutional neural network is trained to obtain the retrained first convolutional neural network; the retrained first convolutional neural network is used as the power consumption data after training Analytical model. 5. The monitoring system for the elderly living alone based on federal multi-source data analysis as claimed in claim 1, characterized in that, water use data analysis; specifically comprising: Using the trained water consumption data analysis model, analyze the real-time water consumption data collected by the water flow sensor, and output the water consumption data analysis results; The water data analysis model after the training; the training step includes: Build a second convolutional neural network; constructing a third training set; the third training set is the historical water consumption data of all the elderly in the whole community with known normal or abnormal labels; The second convolutional neural network is trained by using the third training set, and the trained second convolutional neural network is obtained; constructing a fourth training set; the fourth training set is the historical water consumption data of the elderly to be monitored with known normal or abnormal labels; The fourth training set is used to train the second convolutional neural network after training to obtain the second convolutional neural network after retraining; the second convolutional neural network after retraining is used as the data analysis of water consumption after training Model. 6. The monitoring system for the elderly living alone based on federal multi-source data analysis according to claim 1, wherein the indoor video detection and analysis; specifically comprising: Use the trained indoor video detection model to detect and analyze the indoor video data collected by the cameras installed indoors, and obtain the indoor video detection and analysis results; The indoor video detection model after the training; the specific training process includes: Build a third convolutional neural network; constructing a fifth training set; the fifth training set is the historical indoor activity videos of all elderly people in the whole community with known normal or abnormal labels; The third convolutional neural network is trained by using the fifth training set, and the trained third convolutional neural network is obtained; constructing a sixth training set; the sixth training set is the historical indoor activity videos of the elderly to be monitored with known normal or abnormal labels; The sixth training set is used to train the third convolutional neural network after training, and the third convolutional neural network after retraining is obtained; the third convolutional neural network after retraining is used as the indoor video detection after training. Model. 7. The monitoring system for the elderly living alone based on federal multi-source data analysis according to claim 1, wherein the indoor danger alarm analysis; specifically comprising: Using the trained indoor danger alarm model, the kitchen carbon monoxide concentration collected by the carbon monoxide sensor, the bedroom temperature and humidity collected by the temperature and humidity sensor, and the door lock opening and closing states corresponding to different time points collected by the door lock sensor are detected and analyzed, and the indoor danger is obtained. Alarm analysis results; The trained indoor danger alarm model; the training process includes: Build the fourth convolutional neural network; Constructing a seventh training set; the seventh training set is the carbon monoxide concentration in the kitchen, the temperature and humidity of the bedroom, and the open and closed states of the door locks corresponding to different time points of all the elderly in the whole community with known normal or abnormal labels; Use the seventh training set to train the fourth convolutional neural network, and obtain the fourth convolutional neural network after training; Constructing an eighth training set; the eighth training set is the kitchen carbon monoxide concentration, bedroom temperature and humidity, and door lock open and closed states corresponding to different time points of the elderly to be monitored with known normal or abnormal labels; The eighth training set is used to train the fourth convolutional neural network after training, and the fourth convolutional neural network after retraining is obtained; the fourth convolutional neural network after retraining is used as the indoor danger alarm after training Model. 8. The single-living elderly monitoring system based on federal multi-source data analysis according to claim 1, wherein the controller predicts the alarm level according to the result of the re-analysis, and activates the alarm according to the prediction result of the alarm level Program; specifically includes: Input the results of power consumption data analysis, water data analysis, indoor video detection analysis and indoor danger alarm analysis into the trained warning level classification model, and output the warning level; The trained warning level classification model; the training process includes: Build a support vector machine classifier; Constructing a ninth training set; the ninth training set is the analysis results of electricity consumption data, water data analysis, indoor video detection analysis and indoor danger alarm analysis of known warning level classification labels; The ninth training set is input into the support vector machine classifier, the classifier is trained, the trained support vector machine classifier is obtained, and the trained support vector machine classifier is used as the trained early warning level classification model. 9. The single-living senior citizen monitoring system based on federal multi-source data analysis according to claim 1, wherein the alarm scheme is activated according to the prediction result of the alarm level; comprising: If it is "no risk" level, no alarm; If it is at the level of "potential risk", send a text message reminder to the mobile terminal of the family members of the elderly living alone; If it is "at risk", call the emergency contact directly. 10. A monitoring method for the elderly living alone based on federal multi-source data analysis, characterized in that it includes: The controller, according to the data collected by the electricity meter, the water flow sensor, the camera, the carbon monoxide sensor, the temperature and humidity sensor and the door lock sensor, analyzes the electricity consumption data, the water consumption data analysis, the indoor video detection analysis and the indoor danger alarm analysis; When any alarm situation occurs in the analysis data, the controller controls the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor and door lock sensor to improve the monitoring sensitivity; According to the data collected by the electricity meter, water flow sensor, camera, carbon monoxide sensor, temperature and humidity sensor, and door lock sensor after the monitoring sensitivity is improved, the controller conducts electricity consumption data analysis, water consumption data analysis, indoor video detection analysis and indoor danger alarm again. analyze; The controller predicts the alarm level according to the result of the re-analysis, and starts the alarm scheme according to the prediction result of the alarm level.

Images