Information Management System of Student Laboratory Based on BP
Neural Network
Yunfan Sun
Applied Electronics Department, Shandong Institute Commerce and Technology, Jinan 250103, Shandong, China
Keywords: BP Neural Network, Laboratory Informatization, Management System, Student Laboratory.
Abstract: The laboratory is an important part of cultivating students' practical ability and experimental skills. With the
development of education, the teaching mode of experimental courses has changed from traditional
experimental teaching to today's open experimental teaching. The purpose of this paper is to research
student laboratory information management system based on BP neural network. The investigation and
analysis of the student laboratory information management system based on BP neural network is carried
out, and the key technologies involved in the construction of the system are discussed. Using the advantages
of artificial neural network in data prediction, a three-layer feedforward network model based on BP
algorithm is built, and a framework corresponding to this model is constructed in the system, and the
predicted value of laboratory IGBT devices is verified by simulation results. The results show that the BP
neural network can accurately predict the number of IGBT devices.
1 INTRODUCTION
Laboratory safety in colleges and universities has
become an important part of scientific management
and healthy development in colleges and
universities. With the continuous improvement of
the opening scale of colleges and universities and
the increase of the mobility of examiners, new
challenges have been brought to the informatization
services of laboratories (Wines 2019). At present,
there is a lack of comprehensive assessment
standards for laboratory safety management in
colleges and universities (Jenica 2019). Therefore,
how to accurately and truly evaluate the safety level
of the scientific organization and fair index system
in colleges and universities is a problem that college
administrators should think deeply about (Valencia
2018).
In order to further strengthen the management of
the teaching open laboratory and improve the
utilization rate of equipment, some scholars
proposed a teaching open laboratory management
information system based on video surveillance and
fingerprint access control. The system uses the .NET
programming environment and SQL Server database
system to construct a The software structure mode
combining B/S and C/S. On the basis of remote
video surveillance and fingerprint access control
system, the system realizes online examination
appointment approval, experimental report
submission and modification, online question and
answer, remote video surveillance, fingerprint
access control and other services through learning
and opening the laboratory management website. It
improves the teaching management efficiency of the
open laboratory and provides a good environment
for the cultivation of students' innovative ability
(Johanyák 2019). SA Róański discussed and
demonstrated the necessity of using computer-aided
experiments in the teaching of physics. The benefits
of using computer-aided measurement methods have
been shown. The application of selecting a
measurement console equipped with sensors and
coupled to a computer during the execution and
analysis of experimental results is demonstrated.
The results of three computer-aided experiments are
presented, in which the electromotive force and
internal resistance of the battery, the hysteresis loop
and the Dulong-Petit law are determined (Róański
2020). Therefore, in order to enhance the laboratory
management of the school, it is very necessary to
establish a university experimental learning platform
based on scientific learning management and
advanced information technology, and to establish a
fully operational training laboratory network
362
Sun, Y.
Information Management System of Student Laboratory Based on BP Neural Network.
DOI: 10.5220/0011912200003613
In Proceedings of the 2nd International Conference on New Media Development and Modernized Education (NMDME 2022), pages 362-366
ISBN: 978-989-758-630-9
Copyright
c
 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

management platform (Ummu, Yildiz 2019).
This paper aims to develop a school-based
resource laboratory information management
system. The current situation of the school
laboratory is analyzed. The method and design
scheme of the development system are proposed and
practiced. Make the work of the laboratory truly
efficient, labor-saving and fast. It facilitates school
management, facilitates the daily work and study
arrangements of teachers and students, further
improves the school's informatization management
level, and indirectly supports the development of the
school's new curriculum reform.
2 RESEARCH ON
INFORMATION
MANAGEMENT SYSTEM OF
STUDENT LABORATORY
BASED ON BP NEURAL
NETWORK
2.1 System Design Goals
The laboratory information management system can
integrate and intelligently manage the daily work of
the school's physics, chemistry, biology, and
computer laboratories (Mohammadi 2018). Make
the work of the laboratory truly efficient,
labor-saving and fast. It is convenient for school
management, and also facilitates the daily work and
study arrangements of teachers and students, and
indirectly supports the reform of the new curriculum
(Mcwatt 2018). The design of this system should
achieve the following goals, as shown in Figure 1:
Figure 1. System Design Goals
(1) Friendly user interface: the teachers of this
school are the direct users of the system, the overall
beauty of the system and the ease of use of the
system will directly affect the frequency of teachers'
use of the system and the teachers' overall
evaluation of the system;
(2) Safe system: There are a large number of
school-based materials in the system, all of which
are confidential documents, so the system should
provide confidentiality measures as much as
possible to prevent hacking and data leakage;
(3) Reliable resource management system: The
system should have a laboratory management
function module with complete functions and
convenient operation, and the system administrator
can perform a series of operations on the data
through the module. The system should also have
the function of batch processing, such as batch
uploading files, downloading files, deleting files,
etc.;
(4) Data storage and management: With the
increase of system resources, in order not to bring
too much burden to the system, the system should
have a reasonable data storage scheme, that is,
database storage;
(5) Perfect authority management: The users of
the system include the following groups: system
administrators, physics, chemistry, biology,
computer laboratory administrators, teachers and
students. In order to meet the needs of different
users, different users should have different
permissions;
(6) Reliable user authentication: The system
provides built-in user management functions, which
can be unified by the system administrator to add,
modify, and delete teacher users. At the same time, it
also provides students with online registration
through school card information and birth date.
2.2 BP Neural Network
In this paper, the weight value is calculated by the
questionnaire method to obtain a set of optimal
weight values, and then the laboratory safety
evaluation system model is evaluated. BP neural
network learning includes four processes:
feedforward, error propagation, memory training
and consolidation training .
The learning process of BP neural network
includes two stages of forward estimation and error
propagation (Pitts 2020). First, take the two-layer
BP network as an example, assuming that the input
is p, the input layer has r neurons, the hidden layer
has s1 neurons, and the activation function is F1.
The output plane has a meridian element s2, the
corresponding activation function is F2, the output is
Information Management System of Student Laboratory Based on BP Neural Network
363

A, and the target vector is T.
The network structure layer used in this paper is
three layers, which are the input end, the hidden
layer and the output layer. Since the number of
hidden layers is too large, the training time will be
too long. The number of hidden layers and the
number of nodes is the key consideration. Normalize
the initial value again. Because the initial weight is
randomly given, it is easy to cause network bias. If
the initial value is too large or too small, it will
affect the performance of the algorithm. Therefore,
data with variability is normal and limited to a
certain range of values (Kornilov 2020).
Finally, determine the nonlinear correlation
function, select the appropriate transfer function and
training function, and call the newff command to
create a BP neural network.
3 INVESTIGATION AND
RESEARCH ON
INFORMATION
MANAGEMENT SYSTEM OF
STUDENT LABORATORY
BASED ON BP NEURAL
NETWORK
3.1 System Structure
This article describes the contents of the three layers
and the functions to be implemented:
The bottom layer is the laboratory resource layer.
On the basis of the underlying hardware facilities
and operating platforms, a laboratory resource
library (including large-scale experimental
instruments, small-scale experimental equipment,
experimental drugs, and experimental materials) is
established, which is the basis for constituting
laboratory resources. It is actually necessary to
create a resource library, and at the same time create
the Web server and database server required by the
system.
The management layer (functional layer) is the
core of the entire laboratory information
management system, which is divided into physical
discipline management, chemical discipline
management, biological discipline management and
computer discipline management. On the one hand,
it manages the resources of the bottom layer (data
layer), and on the other hand, it provides a standard
and simple service calling interface for the
presentation layer. Shield the difference and
distribution of the underlying (data layer) resources.
On the basis of the bottom layer (data layer), basic
setting management, fixed asset management,
authority management, etc. are realized.
The presentation layer provides various services
to many different types of users. The presentation
layer provides users with basic setting management,
item storage management, item procurement
management, item requisition management, etc. on
the basis of the management layer. Types such as
laboratory administrators, teachers, students, and
system administrators, users have different
permissions, and provide different service
permissions to meet the needs of maximizing the use
of laboratory resources.
3.2 Realization of Demand Information
Prediction for Experimental
Devices
This section takes IGBT devices as an example for
analysis. Other components are similar to this. The
purchase volume of laboratory IGBT devices from
2018 to 2021 is shown in Table 1.
Table 1. Purchases of IGBT devices from 2018 to 2021
years amount
2018 33
2019 51
2020 66
2021 71
In the model for predicting the purchase volume
of IGBT devices in 2022, a three-layer BP neural
network structure is adopted, and 3 training samples,
1 test sample, 2 input data, 1 output data, and 6
hidden layer nodes are selected. number.
4 ANALYSIS AND RESEARCH OF
STUDENT LABORATORY
INFORMATION
MANAGEMENT SYSTEM
BASED ON BP NEURAL
NETWORK
4.1 System Architecture
The system is mainly composed of three parts:
student system module, teacher system module and
system administrator module. The business logic of
NMDME 2022 - The International Conference on New Media Development and Modernized Education
364

each module will be described in detail below.
(1) Student system module
Browsing the experimental information can
better ensure the integrity of the experimental
information during the experiment reservation
process. If students want to make an appointment for
an experiment, they can see the number of
experiment appointments in time to judge whether
the appointment is successful or if they have more
control over the time.
Under special circumstances, students cannot
participate in the experiment in time, and students
can also cancel the experiment appointment. But it
must be before the experiment starts, otherwise the
system will not display the experiment information
and cannot perform the undo function.
(2) Teacher system module
In the previous laboratory management mode,
teachers can only browse students' experimental
reports after class, but can browse the system
application of the experimental center during the
experiment.
Through the laboratory management system,
teachers can timely feedback the experimental
reports submitted by students. Statistics are based on
the results and procedures of student experiments.
(2) System administrator module
System parameter settings.
data backup.
Experimental information management includes:
experimental items, experimental equipment and
materials, experimental personnel, experimental
funding, and experimental data reporting.
4.2 2022 Procurement Forecast of
Laboratory IGBT Devices Based on
Matlab
This section uses the newff() function in NNbox to
build the neural network model of the IGBT device,
and uses this function to determine the transfer
function, the number of network layers, and the
number of neurons in each layer. The syntax of this
function is:
),,},,...,2,1{],,...,2,1[,( PFBLFBTFTFNTFTFSNSSPRnewffnet =
(1)
In the formula, BTF, BLF, and PF are string
variables, which respectively refer to the name of
the network training function, the name of the
network learning function, and the name of the
network performance function; The Si value
represents the number of neurons in the i-th layer in
the network. TFi represents the transfer function of
layer i in the network. The implementation process
of newff is as follows: after the network structure is
built, the init function is called to initialize the
corresponding thresholds and weights to form a
feedforward network and return the net value
(Attardi 2019).
This section uses batch processing to train the
network, and its corresponding function and syntax
format are:
),,(],[ tpNETtraintrnet =
(2)
In the formula: net represents the revised
network, p represents the input matrix, t represents
the output matrix, NET represents the network to be
trained, and tr represents the training record. The
simulation results are shown in Figure 2:
Figure 2. Best training performance is NaN at epochs 24
-1
-6
-9
-12
-16
-2
-6
-8
-12
-15
0
5
10
15
20
mean squared error (mse)
24 Epochs
Best /10(n) Train /10(n)
Information Management System of Student Laboratory Based on BP Neural Network
365

It can be seen from the figure that when the
number of training is close to 10,000 times, the final
mean square error of the learning training data is
lower than 10
-15
, which is basically zero, indicating
that the established model is very close to the real
situation of IGBT device procurement.
5 CONCLUSIONS
One of the important departments of any university
is the laboratory, and one of the conditions for
measuring the comprehensive strength of the
university is the scale and management of the
laboratory. However, because the management
methods and tools are not very advanced, the level
of management personnel is limited, and the public
training has not been effectively allocated, so it is
difficult to implement effectively. This paper
develops a student laboratory information
management system to improve the efficiency of
laboratory work. There are still many deficiencies in
the design of the student laboratory information
management system, such as: the system does not
automatically generate and print various
questionnaires, statistical reports and so on. There is
still room for further improvement and improvement
for these.
REFERENCES
Attardi S M, Gould D J, Pratt R L, et al. A Data‐Driven
Design: Addressing Student Need for an Anatomy Pre‐
matriculation Experience [J]. The FASEB Journal,
2019, 33(S1):607.6-607.6.
Jenica H, Lisa P, Teresa N. Utilization of Apple iPads in
Student Clinical Rotations to Improve Safety and
Streamline Information Access [J]. American Journal
of Clinical Pathology, 2019(Supplement_1):
S100-S101.
Kornilov V S, Khanina I A. Development of ICT
competence in high school students when teaching
physics using digital laboratories [J]. RUDN Journal
of Informatization in Education, 2020, 17(2):146-152.
Mohammadi A, Afshar P, Asif A, et al. Lung Cancer
Radiomics: Highlights from the IEEE Video and
Image Processing Cup 2018 Student Competition [SP
Competitions] [J]. IEEE Signal Processing Magazine,
2018, 36(1):164-173.
Mcwatt S C, Newton G S, Jadeski L. The Impact of a
Novel Computer-assisted Learning Resource on
Student Learning in Undergraduate Dissection-and
Prosection-based Laboratory Environments [J]. The
FASEB Journal, 2019, 33(S1):17.6-17.6.
Pitts D, Riabov V. The low-budget experimental computer
lab boosts students' research [J]. Journal of Computing
Sciences in Colleges, 2020, 35(8):261-263.
SA Róański. Computer-aided Experiments in Student
Physics Laboratory [J]. Acta Physica Polonica B,
Proceedings Supplement, 2020, 13(4):937-942.
Ummu, Yildiz, Findik, et al. Effect of stoma model based
education on knowledge and skill levels of student
nurses: a quasi-experimental study from Turkey. [J].
JPMA. The Journal of the Pakistan Medical
Association, 2019, 69(10):1496-1500.
Valencia J, Mcwatt S, Jadeski L. Does a Curriculum‐
Targeted, Dissection‐Based Laboratory Workbook
Influence Student Learning Outcomes in
Undergraduate Human Anatomy? [J]. The FASEB
Journal, 2018, 32(S1):lb516-lb516.
Wines K S. WVSOM Anatomy Lab Tour Program: An
Osteopathic Medicine Pipeline With Student Teaching
Opportunities [J]. The Journal of the American
Osteopathic Association, 2019, 119(7):456-463.
ZC Johanyák. Fuzzy rule interpolation based model for
student result prediction[J]. Journal of Intelligent &
Fuzzy Systems, 2019, 36(2):999-1008.
NMDME 2022 - The International Conference on New Media Development and Modernized Education
366