CN114647234B - Medical equipment monitoring method and device based on Internet of things and storage medium - Google Patents
Medical equipment monitoring method and device based on Internet of things and storage medium Download PDFInfo
- Publication number
- CN114647234B CN114647234B CN202210559466.XA CN202210559466A CN114647234B CN 114647234 B CN114647234 B CN 114647234B CN 202210559466 A CN202210559466 A CN 202210559466A CN 114647234 B CN114647234 B CN 114647234B
- Authority
- CN
- China
- Prior art keywords
- fault
- medical equipment
- fault diagnosis
- diagnosis model
- compression
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
- G05B23/0262—Confirmation of fault detection, e.g. extra checks to confirm that a failure has indeed occurred
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24065—Real time diagnostics
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention provides a medical equipment monitoring method, a medical equipment monitoring device and a storage medium based on the Internet of things, wherein the method comprises the following steps: training M fault diagnosis models on a monitoring server to obtain M complete fault diagnosis models, acquiring an identifier and a performance parameter value of each medical device by the monitoring server, acquiring a corresponding complete fault diagnosis model based on the identifier of the medical device, compressing the acquired corresponding complete fault diagnosis model based on the performance parameter value to obtain a compressed fault diagnosis model, and then transmitting the compressed fault diagnosis model to the medical device corresponding to the identifier; when the medical equipment is determined to have a preliminary fault by using the compression fault diagnosis model on the medical equipment based on the real-time acquired state data of the medical equipment, generating a data packet by using first fault information, state data and the identification of the medical equipment output by the compression fault diagnosis model, and sending the data packet to the monitoring server. The calculation intensity of the server is reduced, and the fault diagnosis accuracy is improved.
Description
Technical Field
The invention relates to the technical field of artificial intelligence and the Internet of things, in particular to a medical equipment monitoring method and device based on the Internet of things and a storage medium.
Background
With the progress of medical technology, various medical devices are more and more, how to quickly and accurately diagnose the fault of the medical device is a technical problem, the medical device is adopted to carry out diagnosis based on a diagnosis model in the prior art, and the performance of the medical device is influenced when the diagnosis model is operated in real time due to the low computing capacity of the medical device.
In addition, in the prior art, some medical devices are diagnosed by using a server, and since the medical devices are normally operated for most of the time, the bandwidth of the server is greatly occupied, and the performance of the server is reduced, so that the pressure of the server is urgently required to be reduced.
With the development of artificial intelligence technology, machine learning becomes more and more important, but how to obtain a corresponding compression fault diagnosis model according to the performance of medical equipment has no specific solution.
Disclosure of Invention
The present invention proposes the following technical solutions to address one or more technical defects in the prior art.
A medical device monitoring method based on the Internet of things is disclosed, N medical devices are connected to a monitoring server through a gateway, and the method comprises the following steps:
training, namely training M fault diagnosis models on the monitoring server to obtain M complete fault diagnosis models, wherein each complete fault diagnosis model is specific to one type of medical equipment;
a compression step, wherein the monitoring server acquires the identification and the performance parameter value of each medical device, acquires a corresponding complete fault diagnosis model based on the identification of the medical device, and compresses the acquired corresponding complete fault diagnosis model based on the performance parameter value to obtain a compressed fault diagnosis model;
a sending step, in which the monitoring server sends the compression fault diagnosis model to the medical equipment corresponding to the identifier;
a fault judgment step, namely determining whether the medical equipment has a preliminary fault or not on the medical equipment by using the compression fault diagnosis model based on the real-time acquired state data of the medical equipment, and if so, generating a data packet by using first fault information output by the compression fault diagnosis model, the state data and the identification of the medical equipment and sending the data packet to the monitoring server;
wherein M, N is an integer greater than 1.
Furthermore, after receiving the data packet, the monitoring server obtains first fault information, the state data and the identifier of the medical device from the data packet, the monitoring server obtains corresponding complete fault diagnosis models from the M complete fault diagnosis models based on the identifier of the medical device, the monitoring server uses the complete fault diagnosis models to diagnose the medical device based on the state data to obtain second diagnosis information, and the monitoring server judges whether the medical device has a fault based on the first fault information and the second diagnosis information.
Still further, the operation of the monitoring server determining whether the medical device has a fault based on the first fault information and the second diagnostic information is: if the second diagnosis information indicates that the medical equipment has a fault, the medical equipment has the fault, and warning information is sent to a user; if the second diagnosis information shows that the medical equipment has no fault, and the false alarm times of the compression fault diagnosis model are accumulated; if the second diagnosis information indicates that the medical equipment is suspected to have a fault, calculating a fault index based on the first fault information and the second diagnosis information, if the fault index is larger than a first threshold value, determining that the medical equipment has the fault, otherwise, determining that the fault does not exist, wherein the fault index is calculated in the following way:
wherein, the first and the second end of the pipe are connected with each other,
the index of the fault is represented by,
the information on the first failure is represented by,
the second diagnostic information is represented as a second diagnostic information,
the cosine distance between the first fault information and the second diagnostic information is represented.
Furthermore, when the number of false positives of the compression fault diagnosis model is greater than a second threshold value within a certain period of time, a corresponding complete fault diagnosis model is obtained based on an identifier of the medical equipment corresponding to the compression fault diagnosis model, a compression ratio of the corresponding complete fault diagnosis model is reduced, a first compression fault diagnosis model with higher precision than the compression fault diagnosis model is obtained, and the first compression fault diagnosis model is sent to the medical equipment corresponding to the identifier to replace the compression fault diagnosis model.
Further, the operation of compressing the obtained corresponding complete fault diagnosis model based on the performance parameter values to obtain a compressed fault diagnosis model is as follows: calculating a compression reference value com of the medical device based on the performance parameter values:
wherein, the first and the second end of the pipe are connected with each other,
、
、
、
is a weight value of the weight value,
、
、
、
respectively normalizing the processing capacity, the first-level cache, the second-level cache and the virtual storage space of the ith medical equipment;
compressing the complete fault diagnosis model at a first compression ratio if the compression reference value com is greater than or equal to a third threshold value;
compressing the complete fault diagnosis model at a second compression ratio if the compression reference value com is less than or equal to a fourth threshold value;
compressing the complete fault diagnosis model at a third compression ratio if the compression reference value com is greater than a fourth threshold value and less than a third threshold value;
wherein the first compression ratio is smaller than the third compression ratio, which is smaller than the second compression ratio.
The invention also provides a medical equipment monitoring device based on the Internet of things, wherein N pieces of medical equipment are connected to the monitoring server through the gateway, and the device comprises:
the training unit is used for training M fault diagnosis models on the monitoring server to obtain M complete fault diagnosis models, and each complete fault diagnosis model is specific to one type of medical equipment;
the monitoring server is used for acquiring the identification and the performance parameter value of each piece of medical equipment, acquiring a corresponding complete fault diagnosis model based on the identification of the medical equipment, and compressing the acquired corresponding complete fault diagnosis model based on the performance parameter value to obtain a compressed fault diagnosis model;
the transmitting unit is used for transmitting the compression fault diagnosis model to the medical equipment corresponding to the identifier by the monitoring server;
the fault judgment unit is used for determining whether the medical equipment has a preliminary fault or not on the medical equipment by using the compression fault diagnosis model based on the real-time acquired state data of the medical equipment, and if so, generating a data packet by using first fault information output by the compression fault diagnosis model, the state data and the identification of the medical equipment and sending the data packet to the monitoring server;
wherein M, N is an integer greater than 1.
Furthermore, after receiving the data packet, the monitoring server obtains first fault information, the state data and the identifier of the medical device from the data packet, the monitoring server obtains corresponding complete fault diagnosis models from the M complete fault diagnosis models based on the identifier of the medical device, the monitoring server uses the complete fault diagnosis models to diagnose the medical device based on the state data to obtain second diagnosis information, and the monitoring server judges whether the medical device has a fault based on the first fault information and the second diagnosis information.
Still further, the operation of the monitoring server determining whether the medical device has a fault based on the first fault information and the second diagnostic information is: if the second diagnosis information indicates that the medical equipment has a fault, the medical equipment has the fault, and warning information is sent to a user; if the second diagnosis information shows that the medical equipment has no fault, and the false alarm times of the compression fault diagnosis model are accumulated; if the second diagnosis information indicates that the medical equipment is suspected to have a fault, calculating a fault index based on the first fault information and the second diagnosis information, if the fault index is larger than a first threshold value, determining that the medical equipment has the fault, otherwise, determining that the fault does not exist, wherein the fault index is calculated in the following way:
wherein the content of the first and second substances,
the index of the fault is represented by,
the first failure information is represented by the first failure information,
the second diagnostic information is represented as a second diagnostic information,
the cosine distance between the first fault information and the second diagnostic information is represented.
Furthermore, when the number of false positives of the compression fault diagnosis model is larger than a second threshold within a certain time, a corresponding complete fault diagnosis model is obtained based on an identifier of the medical device corresponding to the compression fault diagnosis model, a compression ratio of the corresponding complete fault diagnosis model is reduced, a first compression fault diagnosis model with higher precision than the compression fault diagnosis model is obtained, and the first compression fault diagnosis model is sent to the medical device corresponding to the identifier to replace the compression fault diagnosis model.
Further, the operation of compressing the obtained corresponding complete fault diagnosis model based on the performance parameter values to obtain a compressed fault diagnosis model is as follows: calculating a compression reference value com of the medical device based on the performance parameter values:
wherein the content of the first and second substances,
、
、
、
is a weight value of the weight value,
、
、
、
respectively the processing capacity, the first level cache, the second level cache and the virtual memory of the ith medical equipmentStoring the normalized value of the space;
compressing the complete fault diagnosis model at a first compression ratio if the compression reference value com is greater than or equal to a third threshold value;
compressing the complete fault diagnosis model at a second compression ratio if the compression reference value com is less than or equal to a fourth threshold value;
compressing the complete fault diagnosis model at a third compression ratio if the compression reference value com is greater than a fourth threshold value and less than a third threshold value;
wherein the first compression ratio is smaller than the third compression ratio, which is smaller than the second compression ratio.
The invention also proposes a computer-readable storage medium having stored thereon computer program code which, when executed by a computer, performs any of the methods described above.
The invention has the technical effects that: the invention discloses a medical equipment monitoring method, a medical equipment monitoring device and a storage medium based on the Internet of things, wherein the method comprises the following steps: a training step S101, training M fault diagnosis models on the monitoring server to obtain M complete fault diagnosis models, wherein each complete fault diagnosis model is specific to one type of medical equipment; a compression step S102, in which the monitoring server acquires the identifier and the performance parameter value of each medical device, acquires a corresponding complete fault diagnosis model based on the identifier of the medical device, and compresses the acquired corresponding complete fault diagnosis model based on the performance parameter value to obtain a compressed fault diagnosis model; a sending step S103, wherein the monitoring server sends the compression fault diagnosis model to the medical equipment corresponding to the identifier; and a fault judgment step S104, determining whether the medical equipment has a preliminary fault or not on the medical equipment by using the compression fault diagnosis model based on the real-time acquired state data of the medical equipment, and if so, generating a data packet by using first fault information output by the compression fault diagnosis model, the state data and the identifier of the medical equipment, and sending the data packet to the monitoring server. In the invention, the medical equipment is used as an edge computing node of the Internet of things, the computing capability of the medical equipment is weaker, and the medical equipment does not have the training capability of a complete diagnosis model, so that the corresponding complete diagnosis model needs to be trained on a monitoring server, namely the complete diagnosis models of various types of medical equipment are stored on the server, then the complete diagnosis model is compressed according to the obtained performance parameters to obtain a compressed fault diagnosis model corresponding to the performance participation of the medical equipment, the compressed fault diagnosis model is preliminarily diagnosed on the medical equipment, when the fault is found by the preliminary diagnosis, a data packet is generated by first fault information output by the compressed fault diagnosis model, state data and the identification of the medical equipment and is sent to the monitoring server for comprehensive judgment after the secondary diagnosis, thereby avoiding that all diagnoses are carried out by using the monitoring server and reducing the computing strength of the server, the state of the medical equipment is judged by combining the complete diagnosis results and the compressed diagnosis results, so that the diagnosis accuracy is improved; in the invention, when the inference result of the complete diagnosis model is suspected to have a fault, a specific mode of calculating a fault index based on the first fault information and the second diagnosis information is established, the cosine distance between the first fault information and the second diagnosis information or a function related to the cosine distance is adopted as a corresponding weight to calculate the fault index, and then whether the medical equipment has the fault is determined based on the size of the fault index, namely, the inference result of the compression model and the complete model is combined for calculation, so that the fault diagnosis accuracy of the medical equipment is improved; in the invention, when the number of times of false alarm of the compression fault diagnosis model exceeds a certain threshold value within a certain time, the compression fault diagnosis model is indicated to have too low precision and is not suitable for fault diagnosis, and the compression ratio can be reduced, so that the obtained compression fault diagnosis model with higher precision is sent to the medical equipment corresponding to the identifier to replace the compression fault diagnosis model, thereby reducing the subsequent number of times of false alarm; according to the invention, the corresponding relation between the performance parameters of the medical equipment and the neural network compression ratio is provided, and the calculation mode can perfectly reflect the performance of the medical equipment as the edge calculation equipment of the Internet of things, so that a diagnosis model with corresponding precision is compressed, and in actual use, the corresponding compression ratio can be adjusted according to the number of misinformation and then recompressed.
Drawings
Other features, objects and advantages of the present application will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, made with reference to the accompanying drawings.
Fig. 1 is a flowchart of a medical device monitoring method based on the internet of things according to an embodiment of the invention.
Fig. 2 is a block diagram of an internet of things-based medical device monitoring apparatus according to an embodiment of the present invention.
Detailed Description
The present application will be described in further detail with reference to the following drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the relevant invention and not restrictive of the invention. It should be noted that, for convenience of description, only the portions related to the related invention are shown in the drawings.
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.
Fig. 1 shows an internet of things-based medical device monitoring method, in which N medical devices are connected to a monitoring server through a gateway, the method including:
a training step S101, training M fault diagnosis models on the monitoring server to obtain M complete fault diagnosis models, wherein each complete fault diagnosis model is specific to one type of medical equipment; the M fault diagnosis models can be CNN (convolutional neural network), DNN (deep neural network), SVM (support vector machine) and the like after training, the training process is to train the neural network based on marked historical data, and in the application, the fault diagnosis models corresponding to various types of medical equipment are trained on a server, namely, each type of medical equipment corresponds to one fault diagnosis model.
A compression step S102, in which the monitoring server acquires the identifier and the performance parameter value of each medical device, acquires a corresponding complete fault diagnosis model based on the identifier of the medical device, and compresses the acquired corresponding complete fault diagnosis model based on the performance parameter value to obtain a compressed fault diagnosis model;
a sending step S103, wherein the monitoring server sends the compression fault diagnosis model to the medical equipment corresponding to the identifier;
a fault judgment step S104, determining whether the medical equipment has a preliminary fault or not on the medical equipment by using the compression fault diagnosis model based on the real-time acquired state data of the medical equipment, and if so, generating a data packet by using first fault information output by the compression fault diagnosis model, the state data and the identification of the medical equipment and sending the data packet to the monitoring server; the data packet is generally sent after being encrypted, the encryption can adopt a symmetric or asymmetric encryption mode, and the encryption mode also adopts a mature encryption algorithm in the prior art. Wherein M, N are integers greater than 1.
In the invention, the medical equipment is used as an edge computing node of the Internet of things, the computing capability of the medical equipment is weaker, and the medical equipment does not have the training capability of a complete diagnosis model, so that the corresponding complete diagnosis model needs to be trained on a monitoring server, namely the complete diagnosis models of various types of medical equipment are stored on the server, then the complete diagnosis model is compressed according to the obtained performance parameters to obtain a compressed fault diagnosis model corresponding to the performance participation of the medical equipment, the compressed fault diagnosis model is preliminarily diagnosed on the medical equipment, when the fault is found by the preliminary diagnosis, a data packet is generated by first fault information output by the compressed fault diagnosis model, state data and the identification of the medical equipment and is sent to the monitoring server for comprehensive judgment after the secondary diagnosis, thereby avoiding that all diagnoses are carried out by using the monitoring server and reducing the computing strength of the server, and the diagnosis result of the complete model and the compressed model is combined to judge the state of the medical equipment, so that the diagnosis accuracy is improved, which is the important invention conception embodiment of the invention.
In a preferred embodiment, after receiving the data packet, the monitoring server obtains first fault information, the state data, and an identifier of the medical device from the data packet, the monitoring server obtains corresponding complete fault diagnosis models from the M complete fault diagnosis models based on the identifier of the medical device, the monitoring server uses the complete fault diagnosis models to diagnose the medical device based on the state data to obtain second diagnosis information, and the monitoring server determines whether the medical device has a fault based on the first fault information and the second diagnosis information.
According to the invention, the state data and the identification of the medical equipment are sent to the monitoring server after the compressed fault diagnosis model is used for judging that the medical equipment has the fault primarily, so that a large amount of normal state data are filtered, the network performance is improved, and the pressure of the monitoring server is reduced.
In a preferred embodiment, the operation of the monitoring server determining whether the medical device has a fault based on the first fault information and the second diagnostic information is to: if the second diagnosis information indicates that the medical equipment has a fault, the medical equipment has the fault, and warning information is sent to a user; if the second diagnosis information shows that the medical equipment has no fault, and the false alarm times of the compression fault diagnosis model are accumulated; if the second diagnosis information indicates that the medical equipment is suspected to have a fault, calculating a fault index based on the first fault information and the second diagnosis information, if the fault index is larger than a first threshold value, determining that the medical equipment has the fault, otherwise, determining that the fault does not exist, wherein the fault index is calculated in the following way:
wherein the content of the first and second substances,
the index of the fault is represented by,
the first failure information is represented by the first failure information,
the second diagnostic information is represented as a second diagnostic information,
the cosine distance between the first fault information and the second diagnostic information is represented.
In the invention, when the inference result of the complete diagnosis model is suspected to have a fault, a specific mode for calculating a fault index based on the first fault information and the second diagnosis information is established, a cosine distance or a function related to the cosine distance between the first fault information and the second diagnosis information is adopted as a corresponding weight to calculate the fault index, and then whether the medical equipment has the fault or not is determined based on the size of the fault index, namely, the calculation is carried out by combining the inference results of the compression model and the complete model, so that the accuracy of fault diagnosis of the medical equipment is improved.
In a preferred embodiment, when the number of false positives of the compression fault diagnosis model is greater than a second threshold within a certain time, a corresponding complete fault diagnosis model is obtained based on an identifier of a medical device corresponding to the compression fault diagnosis model, a compression ratio of the corresponding complete fault diagnosis model is reduced, a first compression fault diagnosis model with higher precision than the compression fault diagnosis model is obtained, and the first compression fault diagnosis model is sent to the medical device corresponding to the identifier to replace the compression fault diagnosis model.
In the present invention, under pressureWhen the false alarm times of the reduced fault diagnosis model exceed a certain threshold value within a certain time, the accuracy of the reduced fault diagnosis model is low, the reduced fault diagnosis model is not suitable for fault diagnosis, the compression ratio can be reduced, for example, the original compression ratio is 50 percent and is reduced to 20 percent, namely the compression ratio is reduced by 20 percent, and the definition of the compression ratio of the invention is that
Therefore, the obtained compression fault diagnosis model with higher precision is sent to the medical equipment corresponding to the identifier to replace the compression fault diagnosis model, so that the subsequent false alarm times are reduced, which is another important invention point of the invention.
In a preferred embodiment, the operation of compressing the obtained corresponding complete fault diagnosis model based on the performance parameter values to obtain a compressed fault diagnosis model is as follows: calculating a compression reference value com of the medical device based on the performance parameter values:
wherein the content of the first and second substances,
、
、
、
is a weight value of the weight value,
、
、
、
respectively normalizing the processing capacity, the first-level cache, the second-level cache and the virtual storage space of the ith medical equipment;
compressing the complete fault diagnosis model at a first compression ratio if the compression reference value com is greater than or equal to a third threshold value; this means that the processing power of the medical device is high and the first compression ratio can be set to be small, for example 20%.
Compressing the complete fault diagnosis model at a second compression ratio if the compression reference value com is less than or equal to a fourth threshold value; this means that the calculation performance of the medical apparatus is low, and the second compression ratio can be set to be large, for example, 60%.
Compressing the complete fault diagnosis model at a third compression ratio if the compression reference value com is greater than a fourth threshold value and less than a third threshold value; this indicates that the medical device is of medium computational performance, and the third compression ratio may be set to an intermediate value, such as 30%. In the present invention, the first compression ratio is smaller than the third compression ratio, which is smaller than the second compression ratio.
In the invention, the corresponding relation between the performance parameters of the medical equipment and the compression ratio of the neural network is provided, and the calculation mode can perfectly reflect the performance of the medical equipment as the edge calculation equipment of the Internet of things, so that a diagnosis model with corresponding precision is compressed, and in actual use, the corresponding compression ratio can be adjusted according to the number of false alarms and then recompressed, which is another important invention point of the invention.
Fig. 2 shows an internet of things-based medical device monitoring apparatus of the present invention, wherein N medical devices are connected to a monitoring server through a gateway, the apparatus includes:
a training unit 201, configured to train M fault diagnosis models on the monitoring server to obtain M complete fault diagnosis models, where each complete fault diagnosis model is for one type of medical equipment; the M fault diagnosis models can be CNN (convolutional neural network), DNN (deep neural network), SVM (support vector machine) and the like after training, the training process is to train the neural network based on marked historical data, and in the application, the fault diagnosis models corresponding to various types of medical equipment are trained on a server, namely, each type of medical equipment corresponds to one fault diagnosis model.
The compression unit 202 is configured to collect the identifier and the performance parameter value of each medical device, acquire a corresponding complete fault diagnosis model based on the identifier of the medical device, and compress the acquired corresponding complete fault diagnosis model based on the performance parameter value to obtain a compressed fault diagnosis model;
the sending unit 203, the monitoring server sends the compression fault diagnosis model to the medical equipment corresponding to the identifier;
the fault judgment unit 204 is used for determining whether the medical equipment has a preliminary fault or not on the medical equipment by using the compression fault diagnosis model based on the real-time acquired state data of the medical equipment, and if so, generating a data packet by using first fault information output by the compression fault diagnosis model, the state data and the identification of the medical equipment and sending the data packet to the monitoring server; the data packet is generally sent after being encrypted, the encryption can adopt a symmetric or asymmetric encryption mode, and the encryption mode also adopts a mature encryption algorithm in the prior art. Wherein M, N are integers greater than 1.
In the invention, the medical equipment is used as an edge computing node of the Internet of things, the computing capability of the medical equipment is weaker, and the medical equipment does not have the training capability of a complete diagnosis model, so that the corresponding complete diagnosis model needs to be trained on a monitoring server, namely the complete diagnosis models of various types of medical equipment are stored on the server, then the complete diagnosis model is compressed according to the obtained performance parameters to obtain a compressed fault diagnosis model corresponding to the performance participation of the medical equipment, the compressed fault diagnosis model is preliminarily diagnosed on the medical equipment, when the fault is found by the preliminary diagnosis, a data packet is generated by first fault information output by the compressed fault diagnosis model, state data and the identification of the medical equipment and is sent to the monitoring server for comprehensive judgment after the secondary diagnosis, thereby avoiding that all diagnoses are carried out by using the monitoring server and reducing the computing strength of the server, and the diagnosis result of the complete model and the compressed model is combined to judge the state of the medical equipment, so that the diagnosis accuracy is improved, which is the important invention conception embodiment of the invention.
In a preferred embodiment, after receiving the data packet, the monitoring server obtains first fault information, the state data, and an identifier of the medical device from the data packet, the monitoring server obtains corresponding complete fault diagnosis models from the M complete fault diagnosis models based on the identifier of the medical device, the monitoring server uses the complete fault diagnosis models to diagnose the medical device based on the state data to obtain second diagnosis information, and the monitoring server determines whether the medical device has a fault based on the first fault information and the second diagnosis information.
According to the invention, the state data and the identification of the medical equipment are sent to the monitoring server after the compressed fault diagnosis model is used for judging that the medical equipment has the fault primarily, so that a large amount of normal state data are filtered, the network performance is improved, and the pressure of the monitoring server is reduced.
In a preferred embodiment, the operation of the monitoring server determining whether the medical device has a fault based on the first fault information and the second diagnostic information is to: if the second diagnosis information indicates that the medical equipment has a fault, the medical equipment has the fault, and warning information is sent to a user; if the second diagnosis information shows that the medical equipment has no fault, and the false alarm times of the compression fault diagnosis model are accumulated; if the second diagnosis information indicates that the medical equipment is suspected to have a fault, calculating a fault index based on the first fault information and the second diagnosis information, if the fault index is larger than a first threshold value, determining that the medical equipment has the fault, otherwise, determining that the fault does not exist, wherein the fault index is calculated in the following way:
wherein the content of the first and second substances,
the index of the fault is represented by,
the information on the first failure is represented by,
the second diagnostic information is represented as a second diagnostic information,
the cosine distance between the first fault information and the second diagnostic information is represented.
In the invention, when the inference result of the complete diagnosis model is suspected to have a fault, a specific mode of calculating a fault index based on the first fault information and the second diagnosis information is established, a cosine distance or a function related to the cosine distance between the first fault information and the second diagnosis information is used as a corresponding weight to calculate the fault index, and then whether the medical equipment has the fault is determined based on the size of the fault index, namely, the calculation is carried out by combining the inference results of the compression model and the complete model, so that the accuracy of the fault diagnosis of the medical equipment is improved.
In a preferred embodiment, when the number of false positives of the compression fault diagnosis model is greater than a second threshold within a certain time, a corresponding complete fault diagnosis model is obtained based on an identifier of a medical device corresponding to the compression fault diagnosis model, a compression ratio of the corresponding complete fault diagnosis model is reduced, a first compression fault diagnosis model with higher precision than the compression fault diagnosis model is obtained, and the first compression fault diagnosis model is sent to the medical device corresponding to the identifier to replace the compression fault diagnosis model.
In the invention, when the number of times of false alarm of the compression fault diagnosis model exceeds a certain threshold value within a certain time, the compression model is indicated to have too low precision and is not suitable for fault diagnosis, the compression ratio can be reduced, for example, the original compression ratio is 50 percent and is reduced to 20 percent, namely, the compression ratio is reduced by 20 percent, and the definition of the compression ratio of the invention is that
Therefore, the obtained compression fault diagnosis model with higher precision is sent to the medical equipment corresponding to the identifier to replace the compression fault diagnosis model, so that the subsequent false alarm times are reduced, which is another important invention point of the invention.
In a preferred embodiment, the operation of compressing the obtained corresponding complete fault diagnosis model based on the performance parameter values to obtain a compressed fault diagnosis model is as follows: calculating a compression reference value com of the medical device based on the performance parameter values:
wherein the content of the first and second substances,
、
、
、
is a weight value of the weight value,
、
、
、
respectively normalizing the processing capacity, the first-level cache, the second-level cache and the virtual storage space of the ith medical equipment;
compressing the complete fault diagnosis model at a first compression ratio if the compression reference value com is greater than or equal to a third threshold value; this means that the processing power of the medical device is high and the first compression ratio can be set to be small, for example 20%.
Compressing the complete fault diagnosis model at a second compression ratio if the compression reference value com is less than or equal to a fourth threshold value; this means that the calculation performance of the medical apparatus is low, and the second compression ratio can be set to be large, for example, 60%.
Compressing the complete fault diagnosis model at a third compression ratio if the compression reference value com is greater than a fourth threshold value and less than a third threshold value; this indicates that the medical device is of medium computational performance, and the third compression ratio may be set to an intermediate value, such as 30%. In the present invention, the first compression ratio is smaller than the third compression ratio, which is smaller than the second compression ratio.
In the invention, the corresponding relation between the performance parameters of the medical equipment and the compression ratio of the neural network is provided, and the calculation mode can perfectly reflect the performance of the medical equipment as the edge calculation equipment of the Internet of things, so that a diagnosis model with corresponding precision is compressed, and in actual use, the corresponding compression ratio can be adjusted according to the number of false alarms and then recompressed, which is another important invention point of the invention.
An embodiment of the present invention provides a computer storage medium, on which a computer program is stored, which when executed by a processor implements the above-mentioned method, and the computer storage medium can be a hard disk, a DVD, a CD, a flash memory, or the like.
For convenience of description, the above devices are described as being divided into various units for separate description. Of course, the functionality of the units may be implemented in one or more software and/or hardware when implementing the present application.
From the above description of the embodiments, it is clear to those skilled in the art that the present application can be implemented by software plus necessary general hardware platform. Based on such understanding, the technical solutions of the present application may be essentially implemented or the portions that contribute to the prior art may be embodied in the form of a software product, which may be stored in a storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the apparatuses described in the embodiments or some portions of the embodiments of the present application.
Finally, it should be noted that: although the present invention has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that: modifications and equivalents may be made thereto without departing from the spirit and scope of the invention and it is intended to cover in the claims the invention as defined in the appended claims.
Claims (6)
1. A medical equipment monitoring method based on the Internet of things is characterized in that N medical equipment are connected to a monitoring server through a gateway, and the method comprises the following steps:
training, namely training M fault diagnosis models on the monitoring server to obtain M complete fault diagnosis models, wherein each complete fault diagnosis model is specific to one type of medical equipment;
a compression step, wherein the monitoring server acquires the identification and the performance parameter value of each medical device, acquires a corresponding complete fault diagnosis model based on the identification of the medical device, and compresses the acquired corresponding complete fault diagnosis model based on the performance parameter value to obtain a compressed fault diagnosis model;
a sending step, in which the monitoring server sends the compression fault diagnosis model to the medical equipment corresponding to the identifier;
a fault judgment step, namely determining whether the medical equipment has a preliminary fault or not on the medical equipment by using the compression fault diagnosis model based on the real-time acquired state data of the medical equipment, and if so, generating a data packet by using first fault information output by the compression fault diagnosis model, the state data and the identification of the medical equipment and sending the data packet to the monitoring server;
wherein M, N is an integer greater than 1;
after receiving the data packet, the monitoring server acquires first fault information, the state data and the identification of the medical equipment from the data packet, the monitoring server acquires corresponding complete fault diagnosis models from the M complete fault diagnosis models based on the identification of the medical equipment, the monitoring server diagnoses the medical equipment based on the state data by using the complete fault diagnosis models to acquire second diagnosis information, and the monitoring server judges whether the medical equipment has faults or not based on the first fault information and the second diagnosis information;
the operation of the monitoring server for judging whether the medical equipment has the fault or not based on the first fault information and the second diagnosis information is as follows: if the second diagnosis information indicates that the medical equipment has a fault, the medical equipment has the fault, and warning information is sent to a user; if the second diagnosis information shows that the medical equipment has no fault, and the false alarm times of the compression fault diagnosis model are accumulated; if the second diagnosis information indicates that the medical equipment is suspected to have a fault, calculating a fault index based on the first fault information and the second diagnosis information, if the fault index is larger than a first threshold value, determining that the medical equipment has the fault, otherwise, determining that the fault does not exist, wherein the fault index is calculated in the following way:
wherein the content of the first and second substances,
the index of the fault is represented by,
the first failure information is represented by the first failure information,
the second diagnostic information is represented as a second diagnostic information,
the cosine distance between the first fault information and the second diagnostic information is represented.
2. The method according to claim 1, wherein when the number of false positives of the compression fault diagnosis model is greater than a second threshold within a certain period of time, a corresponding complete fault diagnosis model is obtained based on an identifier of a medical device corresponding to the compression fault diagnosis model, a compression ratio of the corresponding complete fault diagnosis model is reduced, a first compression fault diagnosis model with higher accuracy than the compression fault diagnosis model is obtained, and the first compression fault diagnosis model is sent to the medical device corresponding to the identifier to replace the compression fault diagnosis model.
3. The method according to claim 2, wherein compressing the obtained corresponding complete fault diagnosis model based on the performance parameter values to obtain a compressed fault diagnosis model comprises: calculating a compression reference value com of the medical device based on the performance parameter values:
wherein the content of the first and second substances,
、
、
、
is a weight value of the weight value,
、
、
、
respectively normalizing the processing capacity, the first-level cache, the second-level cache and the virtual storage space of the ith medical equipment;
compressing the complete fault diagnosis model at a first compression ratio if the compression reference value com is greater than or equal to a third threshold value;
compressing the complete fault diagnosis model at a second compression ratio if the compression reference value com is less than or equal to a fourth threshold value;
compressing the complete fault diagnosis model at a third compression ratio if the compression reference value com is greater than a fourth threshold value and less than a third threshold value;
wherein the first compression ratio is smaller than the third compression ratio, which is smaller than the second compression ratio.
4. A medical equipment monitoring device based on the Internet of things is characterized in that N medical equipment is connected to a monitoring server through a gateway, and the device comprises:
the training unit is used for training M fault diagnosis models on the monitoring server to obtain M complete fault diagnosis models, and each complete fault diagnosis model is specific to one type of medical equipment;
the monitoring server is used for acquiring the identification and the performance parameter value of each piece of medical equipment, acquiring a corresponding complete fault diagnosis model based on the identification of the medical equipment, and compressing the acquired corresponding complete fault diagnosis model based on the performance parameter value to obtain a compressed fault diagnosis model;
the transmitting unit is used for transmitting the compression fault diagnosis model to the medical equipment corresponding to the identifier by the monitoring server;
the fault judgment unit is used for determining whether the medical equipment has a preliminary fault or not on the medical equipment by using the compression fault diagnosis model based on the real-time acquired state data of the medical equipment, and if so, generating a data packet by using first fault information output by the compression fault diagnosis model, the state data and the identification of the medical equipment and sending the data packet to the monitoring server;
wherein M, N is an integer greater than 1;
after receiving the data packet, the monitoring server acquires first fault information, the state data and identification of the medical equipment from the data packet, acquires corresponding complete fault diagnosis models from the M complete fault diagnosis models based on the identification of the medical equipment, diagnoses the medical equipment by using the complete fault diagnosis models based on the state data to acquire second diagnosis information, and judges whether the medical equipment has faults or not based on the first fault information and the second diagnosis information;
the operation of the monitoring server for judging whether the medical equipment has the fault or not based on the first fault information and the second diagnosis information is as follows: if the second diagnosis information indicates that the medical equipment has a fault, the medical equipment has the fault, and warning information is sent to a user; if the second diagnosis information shows that the medical equipment has no fault, and the false alarm times of the compression fault diagnosis model are accumulated; if the second diagnosis information indicates that the medical equipment is suspected to have a fault, calculating a fault index based on the first fault information and the second diagnosis information, if the fault index is larger than a first threshold value, determining that the medical equipment has the fault, otherwise, determining that the fault does not exist, wherein the fault index is calculated in the following way:
wherein the content of the first and second substances,
an index of the failure is indicated,
the first failure information is represented by the first failure information,
the second diagnostic information is represented as a second diagnostic information,
the cosine distance between the first fault information and the second diagnostic information is represented.
5. The apparatus according to claim 4, wherein when the number of false positives of the compression fault diagnosis model is greater than a second threshold within a certain time, a corresponding complete fault diagnosis model is obtained based on an identifier of a medical device corresponding to the compression fault diagnosis model, a compression ratio of the corresponding complete fault diagnosis model is reduced, a first compression fault diagnosis model with higher accuracy than the compression fault diagnosis model is obtained, and the first compression fault diagnosis model is sent to the medical device corresponding to the identifier to replace the compression fault diagnosis model.
6. A computer storage medium, characterized in that a computer program is stored on the computer storage medium, which computer program, when being executed by a processor, carries out the method of any one of claims 1-3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210559466.XA CN114647234B (en) | 2022-05-23 | 2022-05-23 | Medical equipment monitoring method and device based on Internet of things and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210559466.XA CN114647234B (en) | 2022-05-23 | 2022-05-23 | Medical equipment monitoring method and device based on Internet of things and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114647234A CN114647234A (en) | 2022-06-21 |
| CN114647234B true CN114647234B (en) | 2022-08-09 |
Family
ID=81997679
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210559466.XA Active CN114647234B (en) | 2022-05-23 | 2022-05-23 | Medical equipment monitoring method and device based on Internet of things and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114647234B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116723083B (en) * | 2023-08-10 | 2024-01-26 | 山东智拓大数据有限公司 | Cloud server online fault diagnosis method and device |
| CN117115937B (en) * | 2023-10-20 | 2024-01-19 | 湖南半岛医疗科技有限公司 | Equipment running state monitoring method and device, cloud equipment and storage medium |
| CN117176560B (en) * | 2023-11-03 | 2024-01-26 | 山东智云信息科技有限公司 | Monitoring equipment supervision system and method based on Internet of things |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106054858A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | Decision tree classification and fault code classification-based vehicle remote diagnosis and spare part retrieval method |
| CN111340223A (en) * | 2020-02-27 | 2020-06-26 | 上海商汤临港智能科技有限公司 | Neural network compression method, target detection method, driving control method and device |
| CN111739619A (en) * | 2020-06-10 | 2020-10-02 | 深圳市好克医疗仪器股份有限公司 | Medical resource sharing method and device and readable storage medium |
| CN112485652A (en) * | 2020-12-09 | 2021-03-12 | 电子科技大学 | Analog circuit single fault diagnosis method based on improved sine and cosine algorithm |
| CN113836646A (en) * | 2020-06-04 | 2021-12-24 | 北京国电思达科技有限公司 | Fan blade defect intelligent analysis method and system based on deep learning |
Family Cites Families (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108629772B (en) * | 2018-05-08 | 2023-10-03 | 上海商汤智能科技有限公司 | Image processing method and device, computer equipment and computer storage medium |
| CN108646697A (en) * | 2018-07-17 | 2018-10-12 | 河南聚合科技有限公司 | A kind of equipment fault remote diagnosis cloud platform |
| CN109543829A (en) * | 2018-10-15 | 2019-03-29 | 华东计算技术研究所（中国电子科技集团公司第三十二研究所） | Method and system for hybrid deployment of deep learning neural network on terminal and cloud |
| CN109462512A (en) * | 2018-12-25 | 2019-03-12 | 深圳开立生物医疗科技股份有限公司 | Medical equipment failure processing method and system, cloud server and Internet of things system |
| US11651192B2 (en) * | 2019-02-12 | 2023-05-16 | Apple Inc. | Compressed convolutional neural network models |
| CN110133500B (en) * | 2019-05-13 | 2021-08-20 | 中国机械设备工程股份有限公司 | Motor online monitoring and fault precursor diagnosis system and method based on multi-layer architecture |
| CN111007799B (en) * | 2019-12-18 | 2021-04-27 | 宁波财经学院 | Numerical control equipment remote diagnosis system based on neural network |
| CN111523081B (en) * | 2020-05-01 | 2023-09-12 | 西北工业大学 | Aeroengine fault diagnosis method based on enhanced gate control circulating neural network |
| CN111709516B (en) * | 2020-06-09 | 2023-07-28 | 深圳先进技术研究院 | Compression method and compression device, storage medium and equipment of neural network model |
| CN112101767B (en) * | 2020-09-09 | 2023-12-26 | 中国石油大学(北京) | Equipment running state edge cloud fusion diagnosis method and system |
| CN112328420A (en) * | 2020-10-26 | 2021-02-05 | 南京燚麒智能科技有限公司 | Method, device and system for detecting equipment fault |
| CN112364987A (en) * | 2020-12-09 | 2021-02-12 | 中国科学院深圳先进技术研究院 | Training method and training system for convolutional neural network model, computer device and storage medium |
| CN113112086B (en) * | 2021-04-22 | 2022-05-13 | 北京邮电大学 | Intelligent production system based on edge calculation and identification analysis |
| CN113408724A (en) * | 2021-06-17 | 2021-09-17 | 博众精工科技股份有限公司 | Model compression method and device |
| CN114281052A (en) * | 2021-12-30 | 2022-04-05 | 中国兵器装备集团自动化研究所有限公司 | Mechanical equipment running state monitoring device, system and method |
-
2022
- 2022-05-23 CN CN202210559466.XA patent/CN114647234B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106054858A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | Decision tree classification and fault code classification-based vehicle remote diagnosis and spare part retrieval method |
| CN111340223A (en) * | 2020-02-27 | 2020-06-26 | 上海商汤临港智能科技有限公司 | Neural network compression method, target detection method, driving control method and device |
| CN113836646A (en) * | 2020-06-04 | 2021-12-24 | 北京国电思达科技有限公司 | Fan blade defect intelligent analysis method and system based on deep learning |
| CN111739619A (en) * | 2020-06-10 | 2020-10-02 | 深圳市好克医疗仪器股份有限公司 | Medical resource sharing method and device and readable storage medium |
| CN112485652A (en) * | 2020-12-09 | 2021-03-12 | 电子科技大学 | Analog circuit single fault diagnosis method based on improved sine and cosine algorithm |
Non-Patent Citations (1)
| Title |
|---|
| 基于SAE深度学习网络在机床位置相关动态特性研究中的应用;李良杰等;《中国优秀硕士学位论文全文数据库（电子期刊）工程科技I辑》;20200315(第3期);第B022-859页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114647234A (en) | 2022-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114647234B (en) | Medical equipment monitoring method and device based on Internet of things and storage medium | |
| CN111967571B (en) | Abnormality detection method and device based on MHMA | |
| US20230123117A1 (en) | Method and Apparatus for Inspecting Wind Turbine Blade, And Device And Storage Medium Thereof | |
| CN115688018B (en) | Method for monitoring state and diagnosing faults of bearing under multiple working conditions | |
| CN109375609B (en) | Abnormal attack detection method and device | |
| CN110187275B (en) | Method and system for detecting health state of permanent magnet motor | |
| TW202130183A (en) | Device and method of handling anomaly detection | |
| CN114462897B (en) | Comprehensive performance evaluation method and device for highway electromechanical system and storage medium | |
| CN108306997B (en) | Domain name resolution monitoring method and device | |
| CN113392893A (en) | Method, device, storage medium and computer program product for positioning service fault | |
| CN113676343B (en) | Fault source positioning method and device for power communication network | |
| CN112269937B (en) | Method, system and device for calculating user similarity | |
| CN113762967A (en) | Risk information determination method, model training method, device, and program product | |
| CN114676593B (en) | Abnormality detection method and related device for textile equipment | |
| CN116611184A (en) | Fault detection method, device and medium for gear box | |
| CN116776137A (en) | Data processing method and electronic equipment | |
| CN114355234A (en) | Intelligent quality detection method and system for power module | |
| CN118035730B (en) | Equipment fault detection method, electronic equipment and storage medium | |
| CN118094368B (en) | Bearing fault diagnosis method and device based on diffusion model and attention mechanism | |
| CN116993319B (en) | Remote equipment health monitoring method and device based on Internet of things | |
| CN117332352B (en) | Lightning arrester signal defect identification method based on BAM-AlexNet | |
| KR102621106B1 (en) | Apparatus and method for fault diagnosis of ship engines | |
| CN112819083B (en) | Hydraulic pump fault prediction method, device, electronic equipment and storage medium | |
| CN113900860B (en) | Data recovery method for wireless sensor network fault node based on CGAN | |
| CN117094814A (en) | Risk early warning method and system for customer repayment and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |