Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5550-5555
Research Article
Analysis On Industrial Internet Of Things Using Deep Neural Multi-Layer Perceptron
Based Model-Based Engineering
P. Senthilkumar
1, Dr.K.Rajesh
21 Department of Instrumentation and Control Engineering, Kalasalingam Academy of Research and Education,
Krishnankoil, India- 626126 Email: senthilpsk.mail@gmail.com
2 Department of Electrical and Electronics Engineering, Kalasalingam Academy of Research and Education,
Krishnankoil, India-626126 Email: k.rajesh@klu.ac.in
Article History: Received: 10 January 2021; Revised: 12 February 2021; Accepted: 27 March 2021; Published online: 28 April 2021
Abstract: In this paper, an analysis is presented using a Deep Neural Multi-Layer Perceptron based Model-based Engineering (DNMLP-MBE) that implements the industrial workflow in based IIoT. The integrated cloud-based IIoT combines cloud features with open connectivity with IoT. In this research, the validation stages consumes high energy for tracking the reference signal and it requires maximum voltage for the pump. In order to improve the tracking of reference signal with reduced energy and minimum voltage to pump, we use ML algorithm namely Artificial Neural Network (DNMLP-MBE) to optimize the operation in the workflow. The simulation is conducted to verify the benefits associated with Cloud-IIoT integration with MBE.
Keywords: IIoT, Cloud, DNMLP-MBE, ML-MBE 1. Introduction
Cloud capabilities turn industrial automation into process industries [2] in combination with the Internet of Things (IoT) [1]. The IoT services complement evolving and available cloud capacity. Various additional services can increase cloud advantage automation in the smart manufacturing industry.
Design and construction need further collaboration across a variety of enterprises and sectors. Collaborative development models or structures such as a cloud-based resource management system power central the broader range of production items so that IMSs can function efficiently [7]. In the framework of Industry 4.0 [16] - [18], IMS is the foundation for any organization plDNMLP-MBEing to use innovative technology in order to develop more valuable adjustment processes and services [7].
IoT integration with the cloud requires proper transition recently and engineering principles are not necessary for transformation. The implementation of cloud integration imposes stringent restrictions on IoT modules because of their reliability, efficiency and security. Therefore the IoT systems must be installed in conjunction to execute the required tasks, in order to efficiently change the traditional model of automation in the industrial sector.
Recently, some methods of machine learning were primarily adopted [8] – [14] for the industrial cloud and IoT automatic framework, with the help of Deep Neural Multi-Layer Perceptron. Moreover, some engineering model methods [15] are useful for the optimization of cloud-based tasks. This paper analyzes the industrial workflow of Cloud-based IIoT using Deep Neural Multi-Layer Perceptron Model-Based Engineering (DNMLP-MBE).
An in-built cloud-based IIoT integrates cloud functions with open IoT networking. The validation phases use high energy to detect the reference signal and demand full voltage for the pump. We use ML algorithm namely Deep Neural Multi-Layer Perceptron Model-Based Engineering (DNMLP-MBE) to maximize the operation of the workflow to increase the monitoring of the reference signal with reduced energy and low voltage to the pump. This automated workflow solves the optimization routine repeatedly to achieve the necessary performance.
2. Proposed Method
This section discusses a cloud-based workflow architecture for the IIoT sensor, integrating the proposed DNMLP-MBE to maximize process flow through an increased range of reference signals. This enables the pumping operation to be maintained optimally at high voltage. In [15] the collection of the reference voltage is not focused and tackles the task with DNMLP-MBE which preferably finds the maximum voltage for operation. In this case the analysis determines the difference in the present model.
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5550-5555
Research Article
Figure 1: Cloud-IIoT Architecture
The analysis also neglected to include the task schedule information, which currently collects and schedules a DNMLP-MBE template for the rapid acquisition of inbound IoT system signals in accordance with a cloud-based VM. The Cloud-IIoT architecture suggested by DNMLP-MBE is shown in Figure 1.
Model Based Engineering:
The MBE solution would increase abstraction levels and simplify tasks that are susceptible to error and tasks that are intensive. This reduces the implementation costs and improves the exchange of results, reusability and model verification. Therefore, the automation of these considerations is more extensively carried out from the cloud, but with the interaction with various heterogeneity and realms the automation template continues to become complicated. The present study examines a multi-vision modeling of industrial automation systems in the field of cloud-based MBE solutions.
3.2. DNMLP-MBE for reference voltage selection
The model-based architecture is carried out using a workflow to support the implementation of predictive controllers on Cloud-IIoT model-based applications. It is run at four validation phases before deployment, as stated in [15]. The requirement in the model-based design workflow includes the objective feature selected to monitor the
Sen sin g p lan e IoT devices IoT devices IoT devices IoT devices Bursty Traffic Heterogeneous device Data p lan e VMs VMs VMs VMs Congestion or interference Unstable Traffic IIoT Gateway C o n tr o l p lan e Central Controller Artifical Neural Network (DNN) Optimal reference signal selection Task Allocation Plant level controller Pumping operation
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5550-5555
Research Article
with maximal voltage constraints, e.g. maximum voltage a minimal voltage (min.) objective feature is chosen (V).The DNMLP-MBE controller and the control configuration checking procedure was carried out in a simulated environment in this regard (as shown in Figure 2).
Figure 2: DNMLP-MBE Architecture for Cloud-IoT model
Figure 2 shows the architecture of DNMLP-MBE. The treatment units known as neurons are an DNMLP-MBE. The normal neuron structure and behaviour is mirrored by an artificial neuron. Inputs in a neuron and output are available. The neuron has a neuron stimulation function to enable the neuron to activate. The signal is connected to a weight. The key constituents are the weight and input products and the signal power, such that a neuron accepts multiple inputs and only has one output from various sources.
Figure 3: DNMLP Architecture 3. Results and Discussions
This simulation will check the advantages of MBE integration with Cloud-IIoT. The procedure suggested for the test of the effectiveness of the DNMLP-MBE solution is contrasted with the benchmark method. It also evaluates whether the reference signal is monitored for pumping action for decreased time and tension. The report concludes that Cloud-IIoT modelling activities will be analyzed in addition to the analysis, using the assigned tools to gather and acquire cloud and IIoT data and the DNMLP-MBE operation for the optimum cloud-based feedback signal handling.
Following the optimization of the reference signal voltage, Figure 4 displays computing time using the proposed DNMLP-MBE model. The simulation results show that the mission scheduling has increased its ability to reduce the computational time load compared to current MBE methods since the optimization of the reference signal [15]. This indicates that the approach suggested has a lower calculation time than the current state-of-the-art method when planning a job.
Reference signal voltage
IoT data rate Hidden Layer
Output Layer
Input Bandwidth
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5550-5555
Research Article
Figure 4: Computational Time
After optimization of the reference signal voltage by means of the suggested DNMLP-MBE model, Figure 5 displays the costs. The findings of the simulation show that the job scheduling has changed after optimization of the reference signal, in order to reduce the expense of allocating the work scheduled than the current MBE method [15]. This shows that the approach proposed is efficient in planning the challenge at lower costs than the current state-of-the-art method.
Figure 5: Cost ($) for the scheduled task
The pause after the reference signal voltage optimization with the proposed ARN model is seen in Figure 6. Simulated results indicate that the tutorial scheduling has increased the time required to assign the scheduled task and process the task with a reduced time compared to current MBE method [15] after optimization of the reference signal. This demonstrates that the approach introduced is accurate in timing the work with a reduced delay than the current state-of-the-art process.
0.000 2.000 4.000 6.000 8.000 10.000 12.000 14.000 16.000 18.000 20.000 22.000 20 40 60 80 100 C o mp u tai o n al Ti me ( s) Tasks
ANN-MBE Framework MBE Framework
0 5 10 15 20 25 30 35 40 45 50 20 40 60 80 100 C o st Tasks
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5550-5555
Research Article
Figure 6: Delay in scheduled task
Figure 5: Response Time of optimal task allocation task
The reaction times of the reference signal voltage optimization with the proposed DNMLP-MBE model as seen in Figure 6. The findings of the simulation show that the work scheduling has been changed after the referral signal has been optimized by increasing the time the planned task is allocated compared to the current MBE method [15]. This indicates that the proposed system works more effectively with more time than the current state of the art approach to schedule the job.
4. Conclusions
This paper analyzes the industrial workflow of Cloud-based IIoT using DNMLP-MBE. The built-in Cloud-based IIoT integrates cloud functions with open IoT networking. The validation phases use high energy to detect the reference signal and demand full voltage for the pump. A DNMLP-MBE is used to optimize the workflow to enhance the reference signal monitoring with minimal energy and voltage. The automated streamlined workflow addresses optimization restrictions such as delays, costs, calculations and response time to achieve the necessary performance. The validations implement the iterative approach and the implementation modifications can be verified from the performance. This simulation will check the advantages of MBE integration with Cloud-IIoT. The framework proposed is compared to a reference tool for assessing the effectiveness of the proposed approach to deep learning. It also analyzes whether the pumping signal has been tracked with reduced times and low voltage.
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 20 40 60 80 100 C o st Tasks
ANN-MBE Framework MBE Framework
0
2
4
6
8
10
12
14
16
18
20
40
60
80
100
Cost
Tasks
ANN-MBE Framework
MBE Framework
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5550-5555
Research Article
References
1. Wang, C., Bi, Z., & Da Xu, L. (2014). IoT and cloud computing in automation of assembly modeling systems. IEEE Transactions on Industrial Informatics, 10(2), 1426-1434.
2. Gubbi, J., Buyya, R., Marusic, S., & Palaniswami, M. (2013). Internet of Things (IoT): A vision, architectural elements, and future directions. Future generation computer systems, 29(7), 1645-1660. 3. Hossain, M. S., & Muhammad, G. (2016). Cloud-assisted industrial internet of things (iiot)–enabled
framework for health monitoring. Computer Networks, 101, 192-202.
4. Mumtaz, S., Alsohaily, A., Pang, Z., Rayes, A., Tsang, K. F., & Rodriguez, J. (2017). Massive Internet of Things for industrial applications: Addressing wireless IIoT connectivity challenges and ecosystem fragmentation. IEEE Industrial Electronics Magazine, 11(1), 28-33.
5. Kousik, N., Natarajan, Y., Raja, R. A., Kallam, S., Patan, R., & Gandomi, A. H. (2021). Improved salient object detection using hybrid Convolution Recurrent Neural Network. Expert Systems with Applications, 166, 114064.
6. Zhong, R. Y., Huang, G. Q., Lan, S., Dai, Q. Y., Chen, X., & Zhang, T. (2015). A big data approach for logistics trajectory discovery from RFID-enabled production data. International Journal of Production
Economics, 165, 260-272.
7. Natarajan, Y., Kannan, S., & Mohanty, S. N. (2021). Survey of Various Statistical Numerical and Machine Learning Ontological Models on Infectious Disease Ontology. Data Analytics in Bioinformatics: A Machine Learning Perspective, 431-442.
8. Hutter, F., Kotthoff, L., & Vanschoren, J. (2019). Automated Deep Learning. Springer: New York, NY, USA.
9. Géron, A. (2019). Hands-On Deep Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools,
and Techniques to Build Intelligent Systems. O'Reilly Media.
10. Brunton, S. L., Noack, B. R., & Koumoutsakos, P. (2020). Deep Learning for fluid mechanics.
DNMLP-MBEual Review of Fluid Mechanics, 52, 477-508.
11. Schuld, M., & Killoran, N. (2019). Quantum Deep Learning in feature Hilbert spaces. Physical review
letters, 122(4), 040504.
12. Zhang, X. D. (2020). Deep Learning. In A Matrix Algebra Approach to Artificial Intelligence (pp. 223-440). Springer, Singapore.
13. Yang, Q., Liu, Y., Chen, T., & Tong, Y. (2019). Federated Deep Learning: Concept and applications. ACM
Transactions on Intelligent Systems and Technology (TIST), 10(2), 1-19.
14. Wu, Z., Kan, S. J., Lewis, R. D., WittmDNMLP-MBE, B. J., & Arnold, F. H. (2019). Deep Learning-assisted directed protein evolution with combinatorial libraries. Proceedings of the National Academy of
Sciences, 116(18), 8852-8858.
15. Mariappan, L. T., (2020). Analysis On Cardiovascular Disease Classification Using Machine Learning Framework. Solid State Technology, 63(6), 10374-10383.
16. Brilly Sangeetha S., Wilfred Blessing N.R., Adeline Sneha J. (2020) Improving the Training Pattern in Back-Propagation Neural Networks Using Holt-Winters’ Seasonal Method and Gradient Boosting Model. In: Johri P., Verma J., Paul S. (eds) Applications of Deep Learning. Algorithms for Intelligent Systems. Springer, Singapore, pp. 189-198.
17. Raja R, Dhas C (2018) Analysis on improving the response time with PIDSARSA-RAL in ClowdFlows mining platform. EAI Endorsed Trans Energy Web Inf Technol 5(20):1–4
18. Kousik NV, Jayasri S, Daniel A, Rajakumar P (2019) A survey on various load balancing algorithm to improve the task scheduling in cloud computing environment. J Adv Res Dyn Control Syst 11(08):2397– 2406
