Search

CN-122024163-A - Artificial intelligence supervision method and system for incapacitation old man care service

CN122024163ACN 122024163 ACN122024163 ACN 122024163ACN-122024163-A

Abstract

The invention discloses an artificial intelligence supervision method and system for incapacitating old people care services, comprising the steps of converting care monitoring videos into structured care action event objects, preprocessing and collecting the care action event objects according to incapacitating old people identification and care worker identification, carrying out care business process reconstruction and standardized modeling to generate business process log sequences, extracting process supervision indexes and action supervision indexes, constructing an anomaly identification and unified risk assessment model, identifying anomaly event types, obtaining process anomaly scores and action anomaly scores, obtaining unified risk scores and risk grades through a fusion scoring layer, and generating corresponding early warning and rectifying tasks based on the anomaly event types, the unified risk scores and the risk grades. The invention improves the timeliness and traceability of supervision and provides technical support for perfecting the care service system of the disabled old and optimizing the resource allocation.

Inventors

  • LI QINGHUA
  • LI DIANYUAN

Assignees

  • 山东第一医科大学(山东省医学科学院)

Dates

Publication Date
20260512
Application Date
20260204

Claims (10)

  1. 1. An artificial intelligence supervision method for a disabled old man care service is characterized by comprising the following steps: Acquiring a care monitoring video, extracting skeleton key point sequences of a care worker and an incapacitation old person, constructing a space-time skeleton diagram, and identifying a care action unit based on the space-time skeleton diagram to obtain a care action event object; the care service data are obtained, the care service data are preprocessed and collected, and meanwhile, the care action event objects are collected in a correlated mode according to the disabled old man identification and the worker identification; Based on the collected care service data and care action event objects, performing care business process reconstruction and standardized modeling to generate a business process log sequence, and extracting process type supervision indexes and action type supervision indexes; the method comprises the steps of constructing an anomaly identification and unified risk assessment model comprising a flow compliance branch, an action compliance branch and a fusion evaluation layer, identifying an anomaly event type through the flow compliance branch and the action compliance branch, obtaining a flow anomaly score and an action anomaly score, and obtaining a unified risk score and a risk grade through the fusion evaluation layer; Based on the abnormal event type, the unified risk score and the risk grade, corresponding early warning and rectifying tasks are generated.
  2. 2. The method of claim 1, wherein the care action event object comprises an action type, a start-stop time, a duration, an action quality index, a confidence level and an evidence index, wherein the action quality index comprises an action duration, an action completion level, an action stability, a key joint participation distribution and an action subsequence integrity.
  3. 3. The artificial intelligence supervision method for the disabled elderly care service according to claim 1, wherein the care service data comprises care service application and approval records, elderly ability assessment results, service protocols and service plans, care process records and attendance records, patch issue records, organization bed weather and allocation information, mass complaints and work order processing records, and the preprocessing is cleaning, desensitizing and format normalization of the care service data.
  4. 4. The method of claim 1, wherein the process type supervision indexes comprise cross-department collaborative average duration, handling out-of-date rate, capability assessment result and service/subsidy matching degree, caretaking record integrity rate, round allocation fairness, complaint response time and check rate, and the action type supervision indexes comprise action unit coverage rate, action frequency conformity, action duration conformity, action quality conformity rate and service record and action evidence conflict rate.
  5. 5. The method for supervising the artificial intelligence of the care service for the disabled aged as set forth in claim 1, wherein the types of the abnormal events are classified into flow type abnormal events and action type abnormal events, and the flow compliance branches identify that the obtained flow type abnormal events comprise out-of-date handling, node missing, sequence abnormality, abnormal jump and complaint handling are not normal by performing time sequence modeling and consistency analysis on a business flow log sequence.
  6. 6. The artificial intelligence supervision method for the disabled elderly care service according to claim 1, wherein the action compliance branch performs compliance detection and modeling of action coverage, frequency, duration and quality indexes against a care action event object, and the identified action type abnormal events comprise critical action missing, insufficient execution, unqualified quality, evidence and recording conflict.
  7. 7. The artificial intelligence supervision method for the disabled elderly care service according to claim 1, wherein the fusion scoring layer fuses the flow anomaly score and the action anomaly score by means of weighted fusion, gate fusion or rule combination.
  8. 8. An artificial intelligence supervision system for disabled old people care services is characterized by comprising the following modules: the care action event generating module is configured to acquire care monitoring videos, extract skeleton key point sequences of careers and disabled old people, construct a space-time skeleton diagram, identify care action units based on the space-time skeleton diagram and obtain care action event objects; The data collection and preprocessing module is configured to acquire care service data, collect the care service data after preprocessing the care service data, and simultaneously collect the care action event object in a correlation way according to the disabled old man identification and the worker identification; The index extraction module is configured to reconstruct and model the care business process based on the collected care service data and the care action event object, generate a business process log sequence and extract a process supervision index and an action supervision index; the anomaly identification and risk assessment module is configured to construct an anomaly identification and unified risk assessment model comprising a flow compliance branch, an action compliance branch and a fusion assessment layer, identify the type of an anomaly event through the flow compliance branch and the action compliance branch, obtain a flow anomaly score and an action anomaly score, and obtain a unified risk score and a risk grade through the fusion assessment layer; The early warning and rectifying task generating module is configured to generate corresponding early warning and rectifying tasks based on the abnormal event type, the unified risk score and the risk level.
  9. 9. A computer readable storage medium having stored thereon a program, which when executed by a processor performs the steps of an artificial intelligence supervision method for disabled elderly care services according to any of the claims 1-7.
  10. 10. An electronic device comprising a memory, a processor and a program stored on the memory and executable on the processor, wherein the processor, when executing the program, performs the steps in an artificial intelligence supervision method for disabled elderly care services as claimed in any of claims 1 to 7.

Description

Artificial intelligence supervision method and system for incapacitation old man care service Technical Field The invention relates to the technical field of intelligent care and computer intersection, in particular to an artificial intelligent supervision method for the care service of disabled old people. Background The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art. As the population ages and ages continue to deepen, the number of disabled and mentally disabled elderly people continues to increase, and the need for care services for disabled elderly people is becoming increasingly urgent. The existing supervision technology of the care service for the disabled old is mainly based on the traditional supervision mode, and mainly relies on off-line material circulation, manual registration and post spot check to carry out supervision work. Although the partial supervision platform realizes a certain degree of informatization, the reporting of static data and the display of statistical reports are emphasized, and the dominant problems such as overtime, abnormal quantity and the like can be found only through single threshold early warning. In terms of data processing, data of each department and each organization are stored in different business systems in a scattered way, and a unified aggregation, standardization and desensitization processing mechanism is lacked, so that effective data sharing and cooperative utilization are not formed. For supervision of the execution process of the care service, an effective technical means is lacking to extract quantitative and traceable execution evidence, and effective verification of the authenticity, normalization and quality of the service is difficult to realize by taking a care worker filling record or a punch card as a service basis. Therefore, the existing monitoring technology for the care services of the disabled old still has certain defects in practical application: on one hand, the partial supervision mode is mainly on offline material circulation, manual registration and post spot check, and lacks continuous and real-time technical support on the handling progress and timeliness of cross-department collaborative matters, so that the problems of overtime handling, unclear responsibility definition and the like are easy to occur; On the other hand, the ability evaluation results of the disabled old, the maintenance service records, the subsidy issuing information, the mechanism bed position weather, allocation and other data are stored in different service systems in a scattered manner, the data standards are not uniform, the cross-system association analysis difficulty is high, and the overall check on the evaluation results, the service supply and the subsidy issuing consistency is influenced; In addition, the existing part supervision platform focuses on the detection mode based on static rules or thresholds in the aspect of anomaly identification, and the complex business processes involving multiple nodes and multiple roles in the whole process of the care service still lack the capability of fine modeling and comprehensive analysis; meanwhile, in the supervision of the execution process of the care service, the execution evidence is still mainly derived from manual report records or simple attendance information, and a quantifiable and traceable execution evidence system is difficult to form, so that the automatic verification capability of the service authenticity, normalization and quality is limited. Disclosure of Invention In order to solve the problems, the invention provides an artificial intelligent supervision method and an artificial intelligent supervision system for the care service of the disabled aged, which improve the timeliness, fairness, refinement and traceability of supervision and provide technical support for perfecting the care service system of the disabled aged and optimizing the resource allocation. In order to achieve the above purpose, the present invention adopts the following technical scheme: In a first aspect, the present invention provides an artificial intelligence supervision method for a care service of disabled old people, comprising the following steps: Acquiring a care monitoring video, extracting skeleton key point sequences of a care worker and an incapacitation old person, constructing a space-time skeleton diagram, and identifying a care action unit based on the space-time skeleton diagram to obtain a care action event object; the care service data are obtained, the care service data are preprocessed and collected, and meanwhile, the care action event objects are collected in a correlated mode according to the disabled old man identification and the worker identification; Based on the collected care service data and care action event objects, performing care business process reconstruction and standardized modelin