Medical gas intelligent management and control method, system and terminal fusing ai and digital twinning

Abstract

The application relates to a medical gas intelligent management and control method and system fusing AI and digital twinning, and a terminal, and belongs to the technical field of data processing.The medical gas intelligent management and control method comprises the following steps: receiving input information of a gas using equipment; reading an equipment fingerprint library according to the input information, identifying the equipment type and a gas using scene of the gas using equipment; activating a corresponding monitoring protocol according to the equipment type and the gas using scene; acquiring monitoring data of the gas using scene based on the monitoring protocol; analyzing whether the monitoring data is abnormal; if yes, determining an abnormality level, and executing an alarm action according to the abnormality level.The application has the beneficial effect of implementing accurate monitoring for different gas using requirements and use environments, so as to improve monitoring efficiency.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method, system and terminal for intelligent control of medical gases that integrates AI and digital twins. Background Technology

[0002] In the medical field, the safe supply of medical gases is crucial, as it is closely related to patients' lives and health. Medical gases, such as oxygen, nitrogen, and carbon dioxide, play an indispensable role in the prevention, diagnosis, and treatment of patients. For example, oxygen is used in emergency care, respiratory therapy, and treatment of altitude sickness; nitrogen is used in anesthesia and surgery; and carbon dioxide is used in laparoscopic surgery. With the continuous development of medical technology, the application areas of medical gases are constantly expanding, and the requirements for their safe operation and management are also increasing.

[0003] Currently, medical gas system management encompasses multiple aspects, including management responsibilities and systems, operation management of major equipment, use and management of gas cylinders, and safety and emergency management. Hospitals typically establish corresponding management organizations, clearly define management responsibilities and systems, and establish and improve various management systems and operating procedures.

[0004] However, existing methods for managing the safe operation of medical gases largely rely on traditional approaches, lacking precise identification and targeted monitoring of medical gas equipment. Most hospitals use uniform monitoring standards for different medical gas equipment and usage scenarios, which fails to meet diverse needs. For example, the oxygen requirements of operating rooms and general wards differ significantly, yet the same monitoring frequency and standards may be applied, leading to low monitoring efficiency and the inability to promptly detect potential safety hazards. Summary of the Invention

[0005] In order to implement precise monitoring for different gas demand and usage environment and improve monitoring efficiency, this application provides a medical gas intelligent control method, system and terminal that integrates AI and digital twins.

[0006] Firstly, this application provides a method for intelligent control of medical gases that integrates AI and digital twins, employing the following technical solution: A method for intelligent control of medical gases that integrates AI and digital twins includes: Receive input information from gas-consuming equipment; Based on the input information, the device fingerprint database is read to identify the device type and gas usage scenario of the gas-using equipment; Activate the corresponding monitoring protocol based on the device type and the gas usage scenario; Based on the monitoring protocol, monitoring data for the gas consumption scenario is obtained; Analyze whether the monitoring data is abnormal; If so, determine the anomaly level and execute an alarm action based on the anomaly level.

[0007] By adopting the above technical solution, and identifying the type of gas-using equipment and the gas usage scenario, the corresponding monitoring protocol is activated. This enables precise monitoring for different gas usage needs and environments, ensuring a suitable medical gas supply under various conditions. Monitoring data of the gas usage scenario is acquired based on the monitoring protocol, and the data is analyzed in real time for anomalies. Once an anomaly is detected, the anomaly level can be quickly determined and an alarm action can be executed, allowing medical staff and management personnel to be promptly informed of the problem and take appropriate measures, effectively reducing safety risks during the use of medical gases. Intervention in the early stages of abnormal situations prevents the problem from escalating and ensures the safety of patients and medical staff.

[0008] Optionally, the step of analyzing whether the monitoring data is abnormal includes: The first dynamic threshold corresponding to the monitoring data is adaptively calculated based on the gas usage scenario. Obtain the bed occupancy rate for the gas usage scenario; Based on the bed occupancy rate, the first dynamic threshold is adjusted to obtain the second dynamic threshold; The monitoring data is compared with the corresponding second dynamic threshold to determine whether the monitoring data is abnormal.

[0009] By adopting the above technical solution, the first dynamic threshold is adaptively calculated based on the gas usage scenario. This fully considers the characteristics and needs of different gas usage scenarios, making the monitoring more closely aligned with actual conditions, improving the accuracy of anomaly detection, and avoiding the potential for misjudgments or missed detections that may occur with fixed thresholds in complex and ever-changing medical environments. Bed occupancy rate reflects the actual number of patients using medical gases in that scenario and is a crucial factor affecting gas demand. Adjusting the threshold in conjunction with bed occupancy rate allows the monitoring to better reflect the actual load and better balance the supply and demand of medical gases.

[0010] Optionally, the step of adaptively calculating the first dynamic threshold corresponding to the monitoring data based on the gas usage scenario includes: Query the historical database to extract the risk coefficient of the gas usage scenario and the relevant historical monitoring data for the past N days; Calculate the mean and standard deviation of the historical monitoring data; The first dynamic threshold corresponding to the monitoring data is calculated based on the data mean, the standard deviation, and the risk coefficient.

[0011] By employing the aforementioned technical solution and extracting risk coefficients for different gas usage scenarios from historical databases, unique risk factors for various scenarios can be incorporated into the calculation of dynamic thresholds. This allows the thresholds to better adapt to the risk conditions of different scenarios, improving the targeting and effectiveness of monitoring. Historical monitoring data from the past N days is extracted, reflecting the actual gas usage and patterns of the scenario over a period of time, avoiding misjudgments caused by using a uniform fixed threshold. The data mean represents the average level of historical monitoring data, reflecting the general gas usage situation under that scenario; the standard deviation measures the dispersion of the data, reflecting the fluctuation range of gas usage data. These two statistical indicators provide a comprehensive understanding of the distribution characteristics of historical gas usage data. Combining the mean and standard deviation when calculating the first dynamic threshold fully considers the fluctuations in normal gas usage data, ensuring that the threshold covers most normal gas usage situations while effectively identifying data exceeding the normal fluctuation range. By comprehensively considering the statistical characteristics of historical data and the risk factors of the scenario, the limitations of setting thresholds based solely on experience or subjective judgment are avoided, thus enabling more accurate judgment of whether monitoring data is abnormal.

[0012] Optionally, the steps of determining the anomaly level and executing an alarm action based on the anomaly level include: If the anomaly level is level one, the control display screen will output a prompt message. Level one anomaly level means that the difference between the monitored data and the corresponding second dynamic threshold is less than the first difference threshold. If the anomaly level is level two, the alarm will be activated and the duty terminal will display a pop-up window. Level two anomaly level means that the difference between the monitored data and the corresponding second dynamic threshold is not less than the first difference threshold and is less than the second difference threshold. If the anomaly level is level three, the hospital broadcast system will be controlled to broadcast throughout the hospital and the gas supply will be cut off. Level three anomaly level means that the difference between the monitoring data and the corresponding second dynamic threshold is not less than the second difference threshold.

[0013] By adopting the above technical solution, three abnormality levels are defined based on the difference between the monitored data and the second dynamic threshold, accurately reflecting the severity of the abnormality. Level 1 abnormality indicates a small difference, possibly caused by slight fluctuations or accidental factors. In this case, simply displaying a notification on the screen can alert relevant personnel without causing excessive panic or wasting resources. Level 2 abnormality indicates a moderate difference, potentially signifying a certain level of risk. In this case, triggering an alarm and displaying a pop-up message on the duty terminal allows for timely notification of on-duty personnel for further investigation and handling. Level 3 abnormality indicates a large difference, indicating a more serious situation that could significantly impact patient safety and normal hospital operations. In this case, broadcasting a hospital-wide announcement and cutting off the gas supply can quickly raise awareness among all hospital staff and prompt emergency measures to prevent more serious consequences.

[0014] Optionally, the step of analyzing whether the monitoring data is abnormal further includes: If not, then based on the monitoring data, predict the total gas usage time and total gas consumption in the gas usage scenario; Calculate the remaining supply duration based on the total gas consumption and the current gas reserve for the gas consumption scenario; Determine whether the remaining available gas supply time is less than the total gas consumption time; If so, the medical gas dispatch quantity is calculated based on the bed occupancy rate in the gas usage scenario and the difference between the total gas consumption and the current reserve quantity; Based on the gas allocation amount, the hospital's gas usage is allocated, and the allocation information is pushed to the material management terminal.

[0015] By adopting the above technical solution, under the condition that the monitoring data is normal, the total gas usage duration and total gas consumption can be predicted, enabling a more accurate understanding of the medical gas usage situation in the future. The remaining supply duration is calculated based on the predicted total gas consumption and the current reserve, and compared with the predicted total gas consumption duration. This helps hospitals understand in advance whether their medical gas reserves can meet future needs. When it is determined that the remaining supply duration is insufficient, the medical gas allocation amount is calculated based on the bed occupancy rate in the gas usage scenario and the difference between the total gas consumption and the current reserve. Calculating the allocation amount in conjunction with the bed occupancy rate ensures that the allocation is more in line with actual needs, avoiding resource waste due to over-allocation or supply interruptions due to under-allocation. The hospital's gas usage is then allocated according to the calculated allocation amount, and the allocation information is pushed to the materials management terminal. This process achieves dynamic resource allocation, ensuring the rational distribution of medical gas resources among different gas usage scenarios. Staff at the materials management terminal can arrange gas transportation and replenishment in a timely manner based on the allocation information, ensuring a stable gas supply for each gas usage scenario. By forecasting gas demand in advance, calculating and allocating resources, it is possible to effectively avoid interruptions in the supply of medical gases and improve the operational efficiency of hospitals.

[0016] Optionally, the steps following activation of the corresponding monitoring protocol include: Based on the device type and the gas usage scenario, a corresponding virtual security boundary is generated; Real-time acquisition of the spatial location information of the gas-using equipment and its associated environment; The spatial location information is compared with the virtual security boundary in real time; If the gas-using equipment or its associated environment is detected to be outside the virtual safety boundary, a location anomaly alarm is triggered, and the boundary violation information is recorded.

[0017] By adopting the above technical solution, corresponding virtual safety boundaries are generated based on equipment type and gas usage scenario. This fully considers the specific characteristics of different equipment and scenarios, accurately defining the safety range suitable for specific equipment and scenarios. It provides clear boundaries for the safety management of gas-using equipment and its associated environment, avoiding safety accidents caused by improper equipment placement. Once a boundary violation is detected, a location anomaly alarm is immediately triggered, alerting relevant personnel to take measures. This achieves automated detection of abnormal gas equipment positions, eliminating the need for constant manual monitoring of equipment location. The system can automatically complete detection and alarm functions, significantly saving manpower and time costs.

[0018] Optionally, the steps following the alarm action include: The fluctuation curve generated from the monitoring data is compared with the historical fluctuation curves corresponding to historical faults in the historical fault database, and the curve similarity is calculated. The historical fault with the highest curve similarity is used as the predicted fault; Based on the predicted faults, a maintenance work order is generated and pushed to the operation and maintenance terminal.

[0019] By employing the above technical solution, the fluctuation curve generated from the monitoring data is compared with historical fluctuation curves in the historical fault database, and the curve similarity is calculated. The historical fault database stores fluctuation curves corresponding to various faults that have occurred in the past. These curves contain characteristic information about the fault occurrence. By comparing the fluctuation curve of the current monitoring data with historical curves, similarities can be identified from the perspective of data characteristics, thus providing a strong basis for fault prediction. When the current curve is highly similar to a certain historical curve, it means that the current situation may follow a similar fault pattern. Compared with judgment based solely on a single monitoring data or experience, fault prediction based on curve similarity is more objective and accurate, effectively reducing misjudgments and omissions, and providing a more reliable direction for subsequent maintenance work.

[0020] Secondly, this application provides a medical gas intelligent control system that integrates AI and digital twins, employing the following technical solution: A medical gas intelligent control system integrating AI and digital twins includes: The information receiving module is used to receive input information from gas-consuming equipment; The information recognition module is used to read the device fingerprint database based on the input information and identify the device type and gas usage scenario of the gas-using equipment; The data monitoring module is used to activate the corresponding monitoring protocol according to the equipment type and the gas usage scenario, and to obtain the monitoring data of the gas usage scenario based on the monitoring protocol; The data analysis module is used to analyze whether the monitoring data is abnormal; The warning module is used to determine the level of abnormality when the monitored data is abnormal, and to execute an alarm action according to the level of abnormality.

[0021] Thirdly, this application provides a terminal that adopts the following technical solution: A terminal, comprising: The memory contains a medical gas intelligent control program that integrates AI and digital twins; A processor is used to execute a program stored in the memory to implement the steps of the above-described intelligent control method for medical gases that integrates AI and digital twins.

[0022] In summary, this application has at least the following beneficial effects: By identifying the type of gas-using equipment and the gas usage scenario, and activating the corresponding monitoring protocol, precise monitoring can be implemented for different gas usage needs and environments, ensuring a suitable medical gas supply under various conditions. Monitoring data of the gas usage scenario is acquired based on the monitoring protocol, and the data is analyzed in real time for anomalies. Once an anomaly is detected, the anomaly level can be quickly determined and an alarm action can be executed, enabling medical staff and management personnel to be promptly informed of the problem and take appropriate measures, effectively reducing safety risks during the use of medical gases. Intervention in the early stages of abnormal situations prevents the problem from escalating and ensures the safety of patients and medical staff. Attached Figure Description

[0023] Figure 1 This is a first flowchart of an embodiment of the method of this application; Figure 2 This is a second flowchart of an embodiment of the method of this application; Figure 3 This is a third flowchart of an embodiment of the method of this application; Figure 4 This is the fourth flowchart of an embodiment of the method of this application; Figure 5 This is the fifth flowchart of an embodiment of the method of this application; Figure 6 This is the sixth flowchart of an embodiment of the method of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the appendices in the embodiments of the present invention will be described below. Figure 1 - Appendix Figure 6 The technical solutions in the embodiments of the present invention are clearly and completely described herein. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] The first embodiment of this application discloses a method for intelligent control of medical gases that integrates AI and digital twins. (Refer to...) Figure 1 and Figure 2 The intelligent control method for medical gases may include S110-S210: S110, receives input information from gas-consuming equipment; S120: Based on the input information, read the device fingerprint database to identify the device type and gas usage scenario of the gas-using equipment; S130, activate the corresponding monitoring protocol according to the equipment type and gas usage scenario; S140, based on the monitoring protocol, acquires monitoring data of gas consumption scenarios; S150, analyze whether the monitoring data is abnormal; S160, if so, determine the abnormality level and execute the alarm action according to the abnormality level; S170, if not, then based on the monitoring data, predict the total gas usage time and total gas consumption in the gas usage scenario; S180, calculate the remaining supply duration based on the total gas consumption and the current gas reserve for the consumption scenario; S190, determine whether the remaining available gas supply time is less than the total gas consumption time; S200, if so, then calculate the medical gas dispatch volume based on the bed occupancy rate and the difference between the total gas consumption and the current reserve volume in the gas usage scenario; S210, according to the gas dispatch volume, performs resource dispatch for hospital gas use and pushes dispatch information to the material management terminal.

[0026] Specifically, by combining the hospital's existing platform architecture and technical design, a multi-dimensional and intelligent implementation system will be constructed. In the step of receiving input information from gas-using equipment in the S110, relying on the sensing devices and data acquisition gateways of the platform driver layer, intelligent interface modules installed on gas-using equipment (such as ventilators and anesthesia machines) can be used to collect equipment IDs, operating status codes, and initial gas usage parameters in real time. At the same time, the MODBUS-RTU protocol is used to connect to the equipment controller, and the data is transmitted to the big data processing center of the platform layer via Ethernet socket, ensuring the real-time performance and integrity of the input information.

[0027] The identification of equipment type and gas usage scenario for S120 requires the construction of an equipment fingerprint database containing equipment model, manufacturer, and rated parameters at the platform's basic data layer. The fingerprint matching algorithm is called through the equipment and facility management module of the platform's application layer to compare the equipment feature code in the input information with the data in the equipment fingerprint database. At the same time, combined with the department location information (such as ICU, operating room, general ward) from the space management module, the current gas usage scenario is automatically determined. For example, when the equipment model is identified as a ventilator of a certain brand and the location is marked as ICU, the gas usage scenario can be determined to be an intensive care scenario.

[0028] To activate the corresponding monitoring protocol for S130, a scenario-protocol mapping rule base can be established at the application layer: For high-risk scenarios such as the ICU, a multi-parameter monitoring protocol including the OPC UA protocol is activated to connect to the liquid oxygen tank PLC system to obtain liquid level and pressure data. Simultaneously, the MQTT protocol is enabled to connect to pressure transmitters and temperature and humidity sensors within the ward. For ordinary ward scenarios, the lightweight MODBUS-TCP protocol can be used, focusing on monitoring terminal pressure and flow parameters. Furthermore, the monitoring frequencies in the monitoring protocols differ; for example, the monitoring frequency for high-risk scenarios such as the ICU is higher than that for ordinary ward scenarios.

[0029] The S140's monitoring data acquisition is achieved through the sensing layer of the IoT platform. Pressure transmitters, level gauges, and leakage sensors are deployed at the liquid oxygen station and connected to the BA system via the BACnet protocol. At the same time, the medical gas safety operation monitoring module collects oxygen pressure data from various departments and wards. All monitoring data is cleaned by the integrated service in the data governance stage and then uniformly stored in the time-series database of the platform layer.

[0030] In S160, the steps for determining the anomaly level and executing alarm actions based on the anomaly level include: If the anomaly level is level one, the control display screen will output a prompt message. Level one anomaly level means that the difference between the monitored data and the corresponding second dynamic threshold is less than the first difference threshold. If the anomaly level is level two, the alarm will be activated and the duty terminal will display a pop-up window. Level two anomaly level means that the difference between the monitored data and the corresponding second dynamic threshold is not less than the first difference threshold and is less than the second difference threshold. If the anomaly level is level three, the hospital broadcast system will be controlled to broadcast throughout the hospital and the gas supply will be cut off. Level three anomaly level means that the difference between the monitoring data and the corresponding second dynamic threshold is not less than the second difference threshold.

[0031] Specifically, the three-level anomaly response mechanism requires the linkage of multiple system resources: When a Level 1 anomaly occurs (e.g., difference < 0.05 MPa), a yellow warning box flashes on the energy consumption status screen of the corresponding department through the intelligent logistics safety operation management platform at the display layer, and a pending task is generated in the one-stop service module; When a Level 2 anomaly occurs (e.g., 0.05 MPa ≤ difference < 0.1 MPa), the alarm management module at the application layer is immediately triggered to control the activation of the department's audible and visual alarms, and a pop-up message is pushed to the duty terminal via HTTP protocol, including the location of the anomaly, parameter deviation value, and historical trend chart; When a Level 3 anomaly occurs (e.g., difference ≥ 0.1 MPa), the system automatically activates the emergency channel of the hospital broadcast system, sends a hospital-wide broadcast instruction (e.g., "Oxygen supply pressure in the operating room is abnormal, please start the backup manifold immediately") via TCP / IP protocol, and simultaneously calls the equipment control interface to cut off the gas supply valve in the corresponding area. The entire process is displayed in real time on a single visual operation management chart to show the progress of the handling.

[0032] Under normal conditions (S170-S210), the system switches to intelligent scheduling mode. Gas consumption prediction in S170 is based on a predictive analysis model at the platform layer (such as an LSTM neural network model). Features such as gas consumption, bed occupancy rate, and equipment runtime in the same period of the past 7 days are input into the trained predictive analysis model to predict the total gas consumption duration (such as 18 hours) and total gas consumption (such as 800 m³).

[0033] The remaining supply time of S180 needs to be calculated by connecting to the liquid oxygen tank monitoring system's liquid level data (obtained in real time via PLC), combined with the current reserve quantity of the material management system in the basic data layer (e.g., 50 bottles in stock at the central oxygen supply station), and calculated according to the formula "Remaining time = (Current reserve quantity × Single bottle capacity) / Average gas consumption per hour".

[0034] When S190 determines that the remaining time is less than the total gas usage time (e.g., 30 hours are expected but only 25 hours remain), the scheduling quantity calculation module of S200 is activated. It calls the scheduling algorithm of the application layer material management module to generate a scheduling plan: scheduling quantity = difference × (1 + bed occupancy rate × 0.1), ensuring that the scheduling quantity covers the needs of high-load scenarios. In the resource scheduling stage of S210: the system connects to the hospital material management system API via HTTP protocol to generate scheduling work orders (including gas type, quantity, delivery department, and priority), and pushes them to the material management terminal via MQTT protocol. At the same time, the scheduling progress is updated on the comprehensive status screen of the display layer, and the scheduling path is updated in the "Resource Scheduling Map" of the visual operation management platform, realizing intelligent management of the entire process from safety monitoring to resource scheduling.

[0035] Reference Figure 3 S140, the steps for analyzing whether the monitoring data is abnormal may include S310-S340: S310, adaptively calculates the first dynamic threshold corresponding to the monitoring data based on the gas usage scenario; S320, obtain the bed occupancy rate in the gas usage scenario; S330, Based on the bed occupancy rate, the first dynamic threshold is adjusted to obtain the second dynamic threshold; S340 compares the monitoring data with the corresponding second dynamic threshold to determine whether the monitoring data is abnormal.

[0036] Specifically, the dynamic threshold calculation process from S310 to S340 requires a deep integration of historical data and risk assessment models.

[0037] Reference Figure 4 S310, the step of adaptively calculating the first dynamic threshold corresponding to the monitoring data based on the gas usage scenario includes S410-S430: S410, query the historical database to extract the risk coefficient of the gas usage scenario and related historical monitoring data from the past N days; S420, calculates the mean and standard deviation of historical monitoring data; S430: Calculate the first dynamic threshold corresponding to the monitoring data based on the data mean, standard deviation, and risk coefficient.

[0038] Specifically, firstly, the historical database query module at the platform layer extracts historical monitoring data (such as pressure fluctuation curves and average daily gas consumption) for the past 30 days from the historical database for the current gas usage scenario, and calls the statistical analysis tool of the data development module to calculate the data mean and standard deviation; at the same time, combined with the comprehensive situation module at the application layer, the risk coefficient is determined according to the department type (such as the risk coefficient of the operating room is set to 0.8, and that of the general ward is set to 0.5) and the years of equipment operation. For example, the risk coefficient of equipment used for more than 5 years is increased by 20%. Finally, a basic threshold, namely the first dynamic threshold, is generated through a weighted algorithm. The first dynamic threshold = data mean + 1.5 × standard deviation × risk coefficient.

[0039] The bed occupancy rate data of S320 is obtained by connecting to the DB database of the hospital's HIS system and using a web service interface to periodically synchronize the real-time number of beds in each department. The space management module calculates the resource occupancy rate of the current gas usage scenario.

[0040] The threshold adjustment of S330 needs to be dynamically adjusted in conjunction with the tension of medical resources. When the bed occupancy rate exceeds 90% (such as during the peak of related diseases in the ICU), the threshold correction algorithm is called through the energy consumption management module of the application layer. For example, the upper and lower limits of the first dynamic threshold are reduced by 15% to improve monitoring sensitivity and avoid gas supply risks under high load. When the bed occupancy rate is less than 50%, the threshold range can be appropriately widened to reduce false alarms.

[0041] The S340's anomaly detection is achieved through a real-time data comparison engine at the platform layer. It compares the current monitoring data (such as oxygen pressure 0.3MPa) with the second dynamic threshold (such as 0.35MPa-0.55MPa) in milliseconds, and pushes the anomaly results to the alarm module at the application layer in real time.

[0042] Reference Figure 5 The steps following the activation of the corresponding monitoring protocol include S510-S540: S510 generates corresponding virtual safety boundaries based on equipment type and gas usage scenario; S520, real-time acquisition of spatial location information of gas-using equipment and its associated environment; S530 compares spatial location information with virtual security boundaries in real time; S540: If the gas-using equipment or its associated environment is detected to be outside the virtual safety boundary, a location anomaly alarm is triggered and the boundary violation information is recorded.

[0043] Specifically, after activating the corresponding monitoring protocol, the system first generates a virtual safety boundary based on the equipment type (such as ventilators, anesthesia machines, etc.) and the gas usage scenario (such as ICU, operating room, general ward, etc.). This boundary is constructed using 3D modeling technology combined with the hospital's Building Information Modeling (BIM), integrating the equipment's operating range, gas delivery pipeline path, and safety distance parameters (such as a 5-meter fire-free zone around a liquid oxygen tank) to form a dynamically adjustable digital safety area. For example, for anesthesia machines in an operating room, the virtual safety boundary will include the equipment itself, connecting pipelines, and a 1-meter operating space around it, while also incorporating environmental parameters (such as the location of the ventilation system and the coverage area of ​​fire source monitoring sensors).

[0044] Subsequently, the system uses UWB positioning tags (installed on the equipment), infrared sensors, and smart cameras deployed within the gas usage scenario to collect real-time spatial location information of the equipment and its associated environment. The location data is transmitted to the network layer via the IoT platform's perception layer (supporting the MQTT protocol), preprocessed by edge computing nodes, and then uploaded to the platform's big data processing center. The system compares the real-time location information with the virtual security boundary. If it detects that the equipment has exceeded the boundary (e.g., an oxygen cylinder has been moved to an undesignated area) or that the associated environment is abnormal (e.g., personnel have accidentally entered the virtual boundary of a liquid oxygen tank), it immediately triggers a location anomaly alarm. The alarm information is pushed to the maintenance terminal via TCP / IP protocol, and the time of the boundary violation, location coordinates, and associated device ID are simultaneously recorded in the historical database, forming a closed loop for spatial safety monitoring.

[0045] Reference Figure 6 The steps following the execution of the alarm action include S610-S630: S610 compares the fluctuation curve generated from the monitoring data with the historical fluctuation curves corresponding to historical faults in the historical fault database and calculates the curve similarity. S620 uses the historical fault with the highest curve similarity as the predicted fault. The S630 generates maintenance work orders based on predicted faults and pushes them to the maintenance terminal.

[0046] Specifically, after an alarm action is executed (such as a level 2 or higher anomaly), the system automatically extracts the monitoring data generated during the abnormal period to generate a fluctuation curve (such as a pressure drop curve), compares it with typical fault curves stored in the historical fault database (such as pressure reducing valve failure, pipeline leakage), and uses the Dynamic Time Warping (DTW) algorithm to calculate the curve similarity. For example, if the current pressure "sudden rise followed by sudden drop" curve has a similarity of 92% with the historical "pressure reducing valve failure" case, it is determined to be a predicted fault.

[0047] Based on the predicted fault type, the system retrieves the maintenance records of the corresponding equipment from the basic data layer (such as the last pipeline inspection time and material model), and generates standardized maintenance work orders in conjunction with the hospital's equipment and facilities management module. These work orders include the fault location, handling steps (such as closing area valves or replacing seals), required spare parts (such as DN20 stainless steel fittings), and priority (level 3 anomalies correspond to emergency work orders). The work orders are pushed to the operation and maintenance terminal through the one-stop service module, triggering a pending reminder. At the same time, the work order status (pending acceptance, in progress, completed) is recorded at the platform layer, forming a closed-loop management system for the entire process of "monitoring-alarm-prediction-maintenance".

[0048] Based on the above method embodiments, the second embodiment of this application discloses a medical gas intelligent control system integrating AI and digital twins. The medical gas intelligent control system integrating AI and digital twins of this application embodiment can implement any of the above-described methods for intelligent control of medical gases integrating AI and digital twins, and the specific working process of each module in the medical gas intelligent control system integrating AI and digital twins can be referred to the corresponding process in the above method embodiments.

[0049] For ease of understanding, an example is given below: A medical gas intelligent control system integrating AI and digital twins includes: The information receiving module is used to receive input information from gas-consuming equipment; The information recognition module is used to read the device fingerprint database based on the input information and identify the device type and gas usage scenario of the gas-using equipment; The data monitoring module is used to activate the corresponding monitoring protocol according to the equipment type and gas usage scenario, and to obtain monitoring data of the gas usage scenario based on the monitoring protocol; The data analysis module is used to analyze whether the monitoring data is abnormal; The alert module is used to determine the level of abnormality when monitoring data is abnormal, and to execute alarm actions based on the level of abnormality.

[0050] A third embodiment of this application provides a terminal. As one implementation of this terminal, the terminal may include: a memory and a processor; wherein... The memory is used to store the intelligent control program for medical gases that integrates AI and digital twins; The processor is used to execute the program stored in the memory to implement the steps of the above-described intelligent control method for medical gases that integrates AI and digital twins.

[0051] The memory can communicate with the processor via a communication bus, which can be an address bus, a data bus, a control bus, etc.

[0052] Additionally, the memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device.

[0053] Furthermore, the processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0054] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A method for intelligent control of medical gases integrating AI and digital twins, characterized in that, include: Receive input information from gas-consuming equipment; Based on the input information, the device fingerprint database is read to identify the device type and gas usage scenario of the gas-using equipment; Activate the corresponding monitoring protocol based on the device type and the gas usage scenario; Based on the monitoring protocol, monitoring data for the gas consumption scenario is obtained; Analyze whether the monitoring data is abnormal; If so, determine the anomaly level and execute an alarm action based on the anomaly level.

2. The intelligent control method for medical gases integrating AI and digital twins according to claim 1, characterized in that, The steps for analyzing whether the monitoring data is abnormal include: The first dynamic threshold corresponding to the monitoring data is adaptively calculated based on the gas usage scenario. Obtain the bed occupancy rate for the gas usage scenario; Based on the bed occupancy rate, the first dynamic threshold is adjusted to obtain the second dynamic threshold; The monitoring data is compared with the corresponding second dynamic threshold to determine whether the monitoring data is abnormal.

3. The intelligent control method for medical gases integrating AI and digital twins according to claim 2, characterized in that, The step of adaptively calculating the first dynamic threshold corresponding to the monitoring data based on the gas usage scenario includes: Query the historical database to extract the risk coefficient of the gas usage scenario and the relevant historical monitoring data for the past N days; Calculate the mean and standard deviation of the historical monitoring data; The first dynamic threshold corresponding to the monitoring data is calculated based on the data mean, the standard deviation, and the risk coefficient.

4. The intelligent control method for medical gases integrating AI and digital twins according to claim 2, characterized in that, The steps of determining the anomaly level and executing an alarm action based on the anomaly level include: If the anomaly level is level one, the control display screen will output a prompt message. Level one anomaly level means that the difference between the monitored data and the corresponding second dynamic threshold is less than the first difference threshold. If the anomaly level is level two, the alarm will be activated and the duty terminal will display a pop-up window. Level two anomaly level means that the difference between the monitored data and the corresponding second dynamic threshold is not less than the first difference threshold and is less than the second difference threshold. If the anomaly level is level three, the hospital broadcast system will be controlled to broadcast throughout the hospital and the gas supply will be cut off. Level three anomaly level means that the difference between the monitoring data and the corresponding second dynamic threshold is not less than the second difference threshold.

5. The intelligent control method for medical gases integrating AI and digital twins according to claim 1, characterized in that, The steps for analyzing whether the monitoring data is abnormal also include: If not, then based on the monitoring data, predict the total gas usage time and total gas consumption in the gas usage scenario; Calculate the remaining supply duration based on the total gas consumption and the current gas reserve for the gas consumption scenario; Determine whether the remaining available gas supply time is less than the total gas consumption time; If so, the medical gas dispatch quantity is calculated based on the bed occupancy rate in the gas usage scenario and the difference between the total gas consumption and the current reserve quantity; Based on the gas allocation amount, the hospital's gas usage is allocated, and the allocation information is pushed to the material management terminal.

6. The intelligent control method for medical gases integrating AI and digital twins according to claim 1, characterized in that, The steps following activation of the corresponding monitoring protocol include: Based on the device type and the gas usage scenario, a corresponding virtual security boundary is generated; Real-time acquisition of the spatial location information of the gas-using equipment and its associated environment; The spatial location information is compared with the virtual security boundary in real time; If the gas-using equipment or its associated environment is detected to be outside the virtual safety boundary, a location anomaly alarm is triggered, and the boundary violation information is recorded.

7. The intelligent control method for medical gases integrating AI and digital twins according to claim 1, characterized in that, The steps following the execution of the alarm action include: The fluctuation curve generated from the monitoring data is compared with the historical fluctuation curves corresponding to historical faults in the historical fault database, and the curve similarity is calculated. The historical fault with the highest curve similarity is used as the predicted fault; Based on the predicted faults, a maintenance work order is generated and pushed to the operation and maintenance terminal.

8. A medical gas intelligent control system integrating AI and digital twins, characterized in that, The method for intelligent control of medical gases integrating AI and digital twins as described in any one of claims 1-7 includes: The information receiving module is used to receive input information from gas-consuming equipment; The information recognition module is used to read the device fingerprint database based on the input information and identify the device type and gas usage scenario of the gas-using equipment; The data monitoring module is used to activate the corresponding monitoring protocol according to the equipment type and the gas usage scenario, and to obtain the monitoring data of the gas usage scenario based on the monitoring protocol; The data analysis module is used to analyze whether the monitoring data is abnormal; The warning module is used to determine the level of abnormality when the monitored data is abnormal, and to execute an alarm action according to the level of abnormality.

9. A terminal, characterized in that, include: The memory contains a medical gas intelligent control program that integrates AI and digital twins; A processor is configured to execute a program stored in the memory to implement the steps of the intelligent control method for medical gases integrating AI and digital twins as described in any one of claims 1-7.

Images