CN117371933A - Intelligent laboratory management system based on Internet of things - Google Patents

Abstract

The invention discloses an intelligent laboratory management system based on the Internet of things, which belongs to the technical field of Internet of things management and comprises a device management module, a safety monitoring module, a data acquisition module, a reservation module, a scheduling distribution module, an access control module, an energy management module, an alarm notification module, a resource analysis module, a maintenance reminding module, a user platform and a platform optimization module; the method reduces the risk of overfitting, improves the prediction precision, can process diversified data types involved in the resource scheduling of a laboratory, is beneficial to researchers to better know influencing factors of resource utilization, has higher fault tolerance, improves the cache hit rate, can flexibly adjust data in a cache according to an actual access mode, reduces cache pollution and improves the response performance of a user platform.

Description

Intelligent laboratory management system based on Internet of things

Technical Field

The invention relates to the technical field of management of the Internet of things, in particular to an intelligent laboratory management system based on the Internet of things.

Background

In the leading field of science and technology today, laboratories play a key role in promoting innovation and research achievements. From life sciences to physics, from new materials research and development to the medical field, laboratories have been incubators for knowledge discovery and technological advancement. However, with continued depth of research and increased experimental complexity, traditional laboratory management is facing increasing challenges. Traditional laboratory management typically relies on manual intervention, which includes manual maintenance of equipment, manual recording of experimental data, and manual allocation of laboratory resources. This approach is not only prone to errors and inconsistencies, but also wastes a lot of time and resources. Furthermore, laboratory safety and resource utilization are often limited because monitoring and management of laboratory equipment is not comprehensive and timely enough. However, with the development of internet of things, we now have the opportunity to introduce intelligence and automation into laboratory management. The internet of things technology enables various devices and sensors to be connected and communicate in real time, and meanwhile, the rise of artificial intelligence and data analysis technology provides more intelligent tools for data processing and decision making. This means that we can redefine the laboratory management style, improving laboratory efficiency, safety and sustainability.

Through retrieval, chinese patent number CN114971587A discloses an intelligent laboratory management system based on the Internet of things, and the intelligent laboratory management system solves the problems that laboratory experiment equipment, laboratory equipment and laboratory materials are easy to lose and lack, but has low prediction accuracy, cannot process diversified data types involved in laboratory resource scheduling, and is inconvenient for researchers to know influencing factors of resource utilization; in addition, the existing intelligent laboratory management system based on the Internet of things is low in cache hit rate, data in a cache can not be adjusted according to an actual access mode, and response performance of a user platform is reduced.

Disclosure of Invention

The invention aims to solve the defects in the prior art, and provides an intelligent laboratory management system based on the Internet of things.

In order to achieve the above purpose, the present invention adopts the following technical scheme:

an intelligent laboratory management system based on the Internet of things comprises an equipment management module, a safety monitoring module, a data acquisition module, a reservation module, a scheduling and distributing module, an access control module, an energy management module, an alarm notification module, a resource analysis module, a maintenance reminding module, a user platform and a platform optimization module;

the equipment management module is used for connecting laboratory equipment through the Internet of things and automatically detecting the running state of the equipment;

the safety monitoring module is used for monitoring the indoor and outdoor environments of the laboratory through the camera and the sensor;

the data acquisition module is used for acquiring and preprocessing experimental data;

the reservation module is used for a researcher to reserve laboratory equipment and space online;

the dispatching distribution module is used for dispatching equipment and human resources according to an experiment plan and emergency;

the access control module is used for verifying the identity information of the researchers to unlock;

the energy management module is used for monitoring laboratory energy consumption and optimizing equipment use;

the alarm notification module is used for detecting abnormal conditions and sending alarm notification to related personnel;

the resource analysis module is used for collecting and analyzing the use condition of laboratory resources, generating reports and charts, and displaying the utilization rate and bottleneck of the resources;

the maintenance reminding module is used for regularly checking the health condition of the equipment and making a maintenance plan and maintenance reminding;

the user platform is used for a researcher to check and adjust laboratory data;

the platform optimization module is used for optimizing the performance of the user platform.

As a further aspect of the present invention, the sensor used in the safety monitoring module specifically includes a temperature sensor, a humidity sensor, a gas sensor, a flame sensor, a pressure sensor, a motion sensor, an ultraviolet sensor, a radiation detection sensor, a vibration sensor, and a water level sensor.

As a further scheme of the invention, the experimental data preprocessing of the data acquisition module comprises the following specific steps:

step one: analyzing the received experimental data, converting the received experimental data into a processable digital format, removing noise in the data through Gaussian filtering, smoothing the data, calculating standard deviation of the experimental data, and detecting and screening abnormal data according to the calculated standard deviation;

step two: detecting whether repeated data records exist, deleting the repeated data records if the repeated data records exist, detecting missing values existing in each group of data, marking the positions of the missing values in the corresponding data, carrying out statistics and visual analysis on the missing values existing in each group of data to obtain distribution conditions and influence ranges of the missing values, and replacing abnormal values or missing values by the average value or the median of the corresponding K groups of data points found by a KNN algorithm.

As a further scheme of the invention, the specific steps of the dispatching and distributing module equipment and the manpower resource dispatching are as follows:

step I: the scheduling distribution module collects historical decision data from the database, then divides the historical decision data into different state categories and marks the state categories, integrates and generalizes each group of data into a sample data set after standardized processing of the historical decision data, and then divides the sample data set into two groups of feature subsets according to a preset threshold value;

step II: randomly selecting a group of feature subsets, repeatedly performing feature selection and data set segmentation until the depth of a decision tree reaches a preset value, determining the label of a leaf node as the category with the maximum sample number in the node, constructing a complete decision tree through recursion splitting and leaf node label determination, forming a random forest model by using the generated multiple groups of decision trees, and training and optimizing the model;

step III: collecting and processing the latest experimental plan or emergency information, inputting the processed data into the random forest model, starting from the root node of the random forest model, traversing the branches of the tree step by step according to the characteristic conditions of the parameter information until the leaf nodes are reached, taking the labels of the leaf nodes as decision results and outputting the decision results, and scheduling related equipment and human resources according to the decision results.

As a further aspect of the present invention, the energy management module optimizing apparatus uses the following specific steps:

step (1): the energy management module collects laboratory energy use data through sensors and energy metering equipment connected with the Internet of things, pre-processes the collected data, selects characteristics related to energy waste in each group of data, and enables the characteristic scales of each group to be consistent by using a Z-score standardization method;

step (2): collecting various knowledge and information related to laboratory energy consumption from an energy consumption database, classifying, de-duplicating and screening the collected energy consumption knowledge, identifying and extracting entities in the processed energy consumption knowledge through an NLP technology, extracting corresponding attributes of each entity from the related knowledge information, and establishing a relation between the entities to form connection of energy consumption knowledge graphs;

step (3): processing the entity, the attribute and the relation into a corresponding graph structure in a triplet mode, selecting a proper graph database to store and manage the energy consumption knowledge graph, continuously updating and maintaining the energy consumption knowledge graph, then matching the characteristic information with the corresponding entity in the knowledge graph, outputting the mode and trend of energy waste and the equipment or laboratory area of the energy consumption mode consistent with the matching node, and generating a corresponding energy saving strategy.

As a further scheme of the invention, the resource analysis module laboratory resource use condition analysis specifically comprises the following steps:

step 1: the resource analysis module collects the using data of laboratory resources through the Internet of things equipment, integrates the data collected from different data sources into a centralized database, automatically identifies equipment, instruments and facilities in the laboratory through machine learning and image processing technology, and identifies documents and text descriptions of the laboratory resources by using NLP technology;

step 2: preprocessing each group of data, dividing the data into a training set and a testing set, constructing a resource utilization model, carrying out model training through the training set, iterating once in each training period, inputting a group of data into the model in each iteration, calculating a loss value of an output result of the resource utilization model through a cross entropy loss function, and updating weights and parameters through a back propagation algorithm if the loss value does not meet preset requirements;

step 3: the performance of the final trained model is evaluated by using a test set, the trained model is deployed into a related platform, all the collected latest data are input into a resource utilization model to acquire the periodic trend of resource utilization and the utilization rate of each laboratory resource, meanwhile, the allocation of the resources is optimized according to the requirements of different laboratory projects, and the resources are fed back to staff in the form of characters or charts.

As a further scheme of the invention, the specific steps of the performance optimization of the platform optimization module are as follows:

the first step: determining accessed data and data with larger calculation cost in a system according to preset information of researchers, determining pointers pointing to a next node and a last node, and determining a linked list node structure according to the determined data objects and the pointers;

and a second step of: creating an empty linked list, setting the maximum capacity of the linked list according to the memory resource and performance requirements of the system, searching the data in the cache linked list when the data is required to be accessed, moving the data to the head of the linked list if the data exists in the linked list, indicating that the data is used recently, acquiring the data from a database or other data sources if the data is not in the linked list, and adding the data to the head of the linked list;

and a third step of: the method comprises the steps of periodically monitoring the length, cache hit rate and performance index of a linked list, judging the data which are not accessed for the longest time in the linked list based on the latest access time when the cache capacity reaches the upper limit, removing the corresponding data node from the tail of the linked list and releasing resources, simultaneously updating the head pointer of the linked list to a new head node, recording the cache hit rate and the times of elimination operation, and periodically monitoring the system performance.

Compared with the prior art, the invention has the beneficial effects that:

1. according to the method, historical decision data are collected from a database and then divided into different state categories and marked, then the data are divided into two groups of feature subsets according to a preset threshold, a group of feature subsets are randomly selected, feature selection and data set segmentation are repeatedly carried out until the depth of a decision tree reaches a preset value, the label of the leaf node is determined to be the category with the largest sample number in the node, then a complete decision tree is constructed through recursion splitting and leaf node label determination, the generated multiple groups of decision trees form a random forest model, training optimization is carried out on the model, latest experimental plans or emergency information is collected and processed, the processed data are input into the random forest model, branches of the tree are traversed gradually from root nodes of the random forest model according to feature conditions of parameter information until the leaf node is reached, the label of the leaf node is used as a decision result and output, relevant equipment and human resources are scheduled according to the decision result, the overrule is reduced, the prediction accuracy is improved, the diversified data types involved in laboratory resource scheduling can be processed, and the influence on the fault tolerance factor is better known.

2. The invention determines the node structure of the linked list according to the accessed data, the data with larger calculation cost and the pointer structure in the preset information determination system of researchers, creates an empty linked list, searches the data in the buffer linked list when the data is required to be accessed, moves the data to the head of the linked list if the data exists in the linked list, indicates that the data is used recently, acquires the data from a database or other data sources if the data is not in the linked list, adds the data to the head of the linked list, regularly monitors the length of the linked list, the buffer hit rate and the performance index, judges the data which is not accessed for the longest time in the linked list based on the latest access time when the buffer capacity reaches the upper limit, removes and releases resources from the tail of the linked list, updates the head pointer of the linked list to a new head node, records the buffer hit rate and the times of elimination operation, regularly monitors the performance of the system, improves the buffer hit rate, can flexibly adjust the data in the buffer according to the actual access mode, reduces buffer pollution and improves the response performance of a user platform.

Drawings

The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention.

Fig. 1 is a system block diagram of an intelligent laboratory management system based on the internet of things.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments.

Example 1

Referring to fig. 1, an intelligent laboratory management system based on the internet of things comprises a device management module, a security monitoring module, a data acquisition module, a reservation module, a scheduling and distributing module, an access control module, an energy management module, an alarm notification module, a resource analysis module, a maintenance reminding module, a user platform and a platform optimization module.

The equipment management module is used for connecting laboratory equipment through the Internet of things and automatically detecting the running state of the equipment; the safety monitoring module is used for monitoring the indoor and outdoor environments of the laboratory through the camera and the sensor; the data acquisition module is used for acquiring and preprocessing experimental data.

Specifically, the received experimental data are analyzed and converted into a processable digital format, noise in the data is removed through Gaussian filtering, the data is smoothed, standard deviation of the experimental data is calculated, abnormal data are detected and screened out according to the calculated standard deviation, whether repeated data records exist is detected, if the repeated data exist, missing values in all groups of data are detected, positions of all missing values in corresponding data are marked, statistics and visual analysis are conducted on the missing values in all groups of data to obtain distribution conditions and influence ranges of the missing values, and average values or median values of corresponding K groups of data points found through a KNN algorithm replace the abnormal values or the missing values.

In this embodiment, the sensors used in the safety monitoring module include a temperature sensor, a humidity sensor, a gas sensor, a flame sensor, a pressure sensor, a motion sensor, an ultraviolet sensor, a radiation detection sensor, a vibration sensor, and a water level sensor.

The reservation module is used for researchers to reserve laboratory equipment and space online; the scheduling and distributing module is used for scheduling equipment and human resources according to the experiment plan and emergency.

Specifically, the dispatching distribution module collects historical decision data from a database, then divides the historical decision data into different state categories and marks the different state categories, integrates and generalizes each group of data into a sample data set after standardized processing of the historical decision data, then divides the sample data set into two groups of feature subsets according to a preset threshold, randomly selects one group of feature subsets, repeatedly performs feature selection and data set segmentation until the depth of a decision tree reaches a preset value, determines the label of a leaf node as the category with the largest sample number in the node, then determines the label of the leaf node through recursion splitting and the label of the leaf node, constructs a complete decision tree, forms a random forest model by the generated multiple groups of decision trees, trains and optimizes the model, collects and processes the latest experimental plan or emergency information, inputs the processed data into the random forest model, gradually traverses the branches of the tree from the root node of the random forest model according to the feature condition of parameter information until the leaf node is reached, outputs the label of the leaf node as a decision result, and dispatches relevant equipment and manpower resources according to the decision result.

The access control module is used for verifying the identity information of the researchers to unlock; the energy management module is used for monitoring laboratory energy consumption and optimizing equipment use.

Specifically, the energy management module collects laboratory energy use data through sensors and energy metering equipment connected with the Internet of things, pre-processes the collected data, selects characteristics related to energy waste in each group of data, uses a Z-score standardization method to enable each group of characteristic scales to be consistent, collects various knowledge and information related to laboratory energy consumption from an energy consumption database, classifies, de-duplicates and screens the collected energy consumption knowledge, identifies and extracts entities in the processed energy consumption knowledge through an NLP technology, extracts corresponding attributes of each entity from related knowledge information, establishes a relation among the entities, forms connection of energy consumption knowledge graphs, processes the entities, attributes and the relation into a corresponding graph-shaped structure in a triplet mode, selects a proper graph database to store and manage the energy consumption knowledge graphs, continuously updates and maintains the energy consumption knowledge graphs, then matches the characteristic information with the corresponding entities in the knowledge, and outputs a mode of the consistent energy consumption and the energy consumption trend corresponding to the laboratory or energy consumption strategy and generates a corresponding energy consumption strategy area.

Example 2

Referring to fig. 1, an intelligent laboratory management system based on the internet of things comprises a device management module, a security monitoring module, a data acquisition module, a reservation module, a scheduling and distributing module, an access control module, an energy management module, an alarm notification module, a resource analysis module, a maintenance reminding module, a user platform and a platform optimization module.

The alarm notification module is used for detecting abnormal conditions and sending alarm notification to related personnel; the resource analysis module is used for collecting and analyzing the use condition of laboratory resources, generating reports and charts, and displaying the utilization rate and bottleneck of the resources.

Specifically, the resource analysis module collects using data of laboratory resources through the internet of things equipment, integrates the data collected from different data sources into a centralized database, automatically identifies equipment, instruments and facilities in the laboratory through machine learning and image processing technology, identifies documents and text descriptions of the laboratory resources through NLP technology, pre-processes each group of data, divides each group of data into a training set and a testing set, then builds a resource utilization model, carries out model training through the training set, iterates once in each training period, inputs a group of data into the model, calculates the output result loss value of the resource utilization model through a cross entropy loss function in each iteration, updates weights and parameters through a back propagation algorithm if the loss value does not meet preset requirements, evaluates the performance of the finally trained model through the testing set, deploys the trained model into a relevant platform, inputs each group of latest data collected into the resource utilization model to acquire the periodic trend of resource use and the utilization rate of each laboratory resource, simultaneously optimizes the resource allocation according to requirements of different laboratory projects, and feeds back the resource allocation to a working staff or a graph form.

The maintenance reminding module is used for regularly checking the health condition of the equipment and making a maintenance plan and maintenance reminding; the user platform is used for a researcher to check and adjust laboratory data; the platform optimization module is used for optimizing the performance of the user platform.

Specifically, according to preset information of researchers, determining accessed data and data with larger calculation cost in a system, determining pointers pointing to a next node and a last node, determining a node structure of the linked list according to determined data objects and the pointers, creating an empty linked list, simultaneously according to memory resources and performance requirements of the system, setting the maximum capacity of the linked list, searching the data in a cache linked list, if the data exists in the linked list, moving the data to the head of the linked list to indicate that the data is recently used, if the data is not in the linked list, acquiring the data from a database or other data sources, adding the data to the head of the linked list, periodically monitoring the length, cache hit rate and performance index of the linked list, judging the data which is not accessed for the longest time in the linked list based on the latest access time when the cache capacity reaches the upper limit, removing and releasing resources from the tail of the linked list corresponding data nodes, simultaneously updating the head pointer of the linked list to the new head node, recording the cache rate and the number of times of elimination operations, and periodically monitoring the performance of the system.

Claims (7)

1. The intelligent laboratory management system based on the Internet of things is characterized by comprising an equipment management module, a safety monitoring module, a data acquisition module, a reservation module, a scheduling and distributing module, an access control module, an energy management module, an alarm notification module, a resource analysis module, a maintenance reminding module, a user platform and a platform optimization module;

the equipment management module is used for connecting laboratory equipment through the Internet of things and automatically detecting the running state of the equipment;

the safety monitoring module is used for monitoring the indoor and outdoor environments of the laboratory through the camera and the sensor;

the data acquisition module is used for acquiring and preprocessing experimental data;

the reservation module is used for a researcher to reserve laboratory equipment and space online;

the dispatching distribution module is used for dispatching equipment and human resources according to an experiment plan and emergency;

the access control module is used for verifying the identity information of the researchers to unlock;

the energy management module is used for monitoring laboratory energy consumption and optimizing equipment use;

the alarm notification module is used for detecting abnormal conditions and sending alarm notification to related personnel;

the resource analysis module is used for collecting and analyzing the use condition of laboratory resources, generating reports and charts, and displaying the utilization rate and bottleneck of the resources;

the maintenance reminding module is used for regularly checking the health condition of the equipment and making a maintenance plan and maintenance reminding;

the user platform is used for a researcher to check and adjust laboratory data;

the platform optimization module is used for optimizing the performance of the user platform.

2. The intelligent laboratory management system based on the internet of things according to claim 1, wherein the sensor used by the safety monitoring module comprises a temperature sensor, a humidity sensor, a gas sensor, a flame sensor, a pressure sensor, a motion sensor, an ultraviolet sensor, a radiation detection sensor, a vibration sensor and a water level sensor.

3. The intelligent laboratory management system based on the internet of things according to claim 1, wherein the data acquisition module experimental data preprocessing specifically comprises the following steps:

step one: analyzing the received experimental data, converting the received experimental data into a processable digital format, removing noise in the data through Gaussian filtering, smoothing the data, calculating standard deviation of the experimental data, and detecting and screening abnormal data according to the calculated standard deviation;

step two: detecting whether repeated data records exist, deleting the repeated data records if the repeated data records exist, detecting missing values existing in each group of data, marking the positions of the missing values in the corresponding data, carrying out statistics and visual analysis on the missing values existing in each group of data to obtain distribution conditions and influence ranges of the missing values, and replacing abnormal values or missing values by the average value or the median of the corresponding K groups of data points found by a KNN algorithm.

4. The intelligent laboratory management system based on the internet of things according to claim 3, wherein the scheduling and distributing module device and the human resource scheduling specifically comprises the following steps:

step I: the scheduling distribution module collects historical decision data from the database, then divides the historical decision data into different state categories and marks the state categories, integrates and generalizes each group of data into a sample data set after standardized processing of the historical decision data, and then divides the sample data set into two groups of feature subsets according to a preset threshold value;

step II: randomly selecting a group of feature subsets, repeatedly performing feature selection and data set segmentation until the depth of a decision tree reaches a preset value, determining the label of a leaf node as the category with the maximum sample number in the node, constructing a complete decision tree through recursion splitting and leaf node label determination, forming a random forest model by using the generated multiple groups of decision trees, and training and optimizing the model;

step III: collecting and processing the latest experimental plan or emergency information, inputting the processed data into the random forest model, starting from the root node of the random forest model, traversing the branches of the tree step by step according to the characteristic conditions of the parameter information until the leaf nodes are reached, taking the labels of the leaf nodes as decision results and outputting the decision results, and scheduling related equipment and human resources according to the decision results.

5. The intelligent laboratory management system based on the internet of things according to claim 1, wherein the energy management module optimizing device uses the following specific steps:

step (1): the energy management module collects laboratory energy use data through sensors and energy metering equipment connected with the Internet of things, pre-processes the collected data, selects characteristics related to energy waste in each group of data, and enables the characteristic scales of each group to be consistent by using a Z-score standardization method;

step (2): collecting various knowledge and information related to laboratory energy consumption from an energy consumption database, classifying, de-duplicating and screening the collected energy consumption knowledge, identifying and extracting entities in the processed energy consumption knowledge through an NLP technology, extracting corresponding attributes of each entity from the related knowledge information, and establishing a relation between the entities to form connection of energy consumption knowledge graphs;

step (3): processing the entity, the attribute and the relation into a corresponding graph structure in a triplet mode, selecting a proper graph database to store and manage the energy consumption knowledge graph, continuously updating and maintaining the energy consumption knowledge graph, then matching the characteristic information with the corresponding entity in the knowledge graph, outputting the mode and trend of energy waste and the equipment or laboratory area of the energy consumption mode consistent with the matching node, and generating a corresponding energy saving strategy.

6. The intelligent laboratory management system based on the internet of things according to claim 5, wherein the resource analysis module laboratory resource usage analysis specifically comprises the following steps:

step 1: the resource analysis module collects the using data of laboratory resources through the Internet of things equipment, integrates the data collected from different data sources into a centralized database, automatically identifies equipment, instruments and facilities in the laboratory through machine learning and image processing technology, and identifies documents and text descriptions of the laboratory resources by using NLP technology;

step 2: preprocessing each group of data, dividing the data into a training set and a testing set, constructing a resource utilization model, carrying out model training through the training set, iterating once in each training period, inputting a group of data into the model in each iteration, calculating a loss value of an output result of the resource utilization model through a cross entropy loss function, and updating weights and parameters through a back propagation algorithm if the loss value does not meet preset requirements;

step 3: the performance of the final trained model is evaluated by using a test set, the trained model is deployed into a related platform, all the collected latest data are input into a resource utilization model to acquire the periodic trend of resource utilization and the utilization rate of each laboratory resource, meanwhile, the allocation of the resources is optimized according to the requirements of different laboratory projects, and the resources are fed back to staff in the form of characters or charts.

7. The intelligent laboratory management system based on the internet of things according to claim 1, wherein the specific steps of the performance optimization of the platform optimization module are as follows:

the first step: determining accessed data and data with larger calculation cost in a system according to preset information of researchers, determining pointers pointing to a next node and a last node, and determining a linked list node structure according to the determined data objects and the pointers;

and a second step of: creating an empty linked list, setting the maximum capacity of the linked list according to the memory resource and performance requirements of the system, searching the data in the cache linked list when the data is required to be accessed, moving the data to the head of the linked list if the data exists in the linked list, indicating that the data is used recently, acquiring the data from a database or other data sources if the data is not in the linked list, and adding the data to the head of the linked list;

and a third step of: the method comprises the steps of periodically monitoring the length, cache hit rate and performance index of a linked list, judging the data which are not accessed for the longest time in the linked list based on the latest access time when the cache capacity reaches the upper limit, removing the corresponding data node from the tail of the linked list and releasing resources, simultaneously updating the head pointer of the linked list to a new head node, recording the cache hit rate and the times of elimination operation, and periodically monitoring the system performance.