Artificial Intelligence and Chemical Industry

Mid Seminar Report

Submitted in partial fulfilment of the requirement of

B.Tech in Chemical Engineering

by

Arpit Patel

16BCH038

Under the guidance of

Dr. Pravin Kodgire

School of technology

Pandit Deendayal Petroleum University

Gandhinagar – 382007. Gujarat – INDIA

Oct-2019
 Approval sheet

The report entitled “Artificial Intelligence And Chemical Industry” by Arpit Patel is

recommended to be evaluated by the faculty panel.

Supervisor

Dr. Pravin Kodgire

ii
 Student Declaration

I Arpit Patel hereby declare that this written submission represents my ideas in my own

words and where others’ idea or words have been included, I have adequately cited and

referenced the original sources. I also declare that I have adhered to all principles of academic

honestly and integrity and have not misrepresented or fabricated or falsified any idea / data /

fact /source in my submission. I understand that any violation of the above will be cause for

disciplinary action by the PANDIT DEENDAYAL PETROLEUM UNIVERSITY and can

also evoke penal action from the sources which have thus not been properly cited or from

whim proper permission has not been taken when needed.

__________________

Arpit Patel

16BCH038

Date: 16/10/2019

iii
Abstract:

Solving chemical engineering problems due to the highly nonlinear behavior of chemical

processes is often impossible or very difficult. Today, artificial intelligence (AI) techniques

are becoming useful due to simple implementation, easy designing, generality, robustness and

flexibility. The AI includes various branches, namely, artificial neural network, fuzzy logic,

genetic algorithm. They have been widely used in various applications of the chemical

engineering field including modeling, process control, classification, fault detection and

diagnosis. In this chapter, the capabilities of AI are investigated in various chemical

engineering fields.

iv
 Table of content

Chapter Title Page no.

2 Application 2

3 Methods 4

Artificial neural network

Fuzzy logic

Genetic algorithm

4 Data processing and model development 6

5 Application of AI techniques in chemical engineering 8

..

v
List of tables:

No. Title Page no.

List of figures:

No. Title Page no.

vi
1) Introduction to Artificial Intelligence:

Artificial Intelligence is an approach to make a computer, a robot, or a product to think how

smart human think. AI is a study of how human brain think, learn, decide and work, when it

tries to solve problems. And finally this study outputs intelligent software systems. The aim of

AI is to improve computer functions which are related to human knowledge, for example,

reasoning, learning, and problem-solving.

After the 4.0 evolution in industries, applications of Artificial intelligence (AI) in chemical

engineering have increased dramatically recently. AI is a field of computer science that can

simulate characteristics of human intelligence and human sensory capabilities. AI systems

can provide superior solutions over classical systems due to their heuristic and intelligent

nature. For example, it is too difficult to use classical systems to get global optima for the

assembly line balancing problem, which can be easily achieved by the use of the AI.

Conventional computers lack the ability to learn, and this restricts them to operate only under

the conditions for which they are programmed. Applications of AI in the chemical

engineering field including process such as modelling, optimization, process control, fault

detection and diagnosis. [1]

1
2) Application of AI:

AI is used in wide range of processes in chemical industry. The process control strategies

have been developed to improve the performance of the process, reduce energy consumption

and ensure high safety and environmental goals. The conventional controllers cannot show

satisfactory responses in many industrial chemical processes with high nonlinear dynamics

and parameter uncertainties, whereas AI approaches can be effectively controlled for a

number of complex and nonlinear processes [2].

AI is now used in all industry from farming to refinery. It is used in weather prediction and in

water purity measurement. AI works on basis of available data so it is believed that data is

new oil.

There are two major types of modelling approaches in chemical engineering, namely,

mechanistic (white box, first principle) and AI-based approach like ANN and fuzzy logic

methods. In the mechanistic approach, fundamental physical and chemical laws, such as

conservation laws, construct the basis of the model. This approach contains algebraic and

differential equations which involve mass, energy and momentum balances. Due to the large

number of variables affecting the process behaviour and complex mathematical equations

governing the system, many chemical processes are nonlinear and complicated.

Consequently, it is hard and sometimes even impossible to present them by mechanistic

models. Even if such a model has been developed, it might be impractical to solve or identify

its parameters. Moreover, a mechanistic model needs detailed knowledge and a lot of skill

and ingenuity to incorporate the basic phenomena of the process in the model. Difficulties

can arise from poor knowledge [2]. In some cases, considering some assumptions such as

physical properties’ constancy, ideality of gas phase and linearization of the nonlinear

2
equations of the model is inevitable, which all impose limitations on the model leading to the

reduction of the model’s robustness [4].

On the contrary, AI-based techniques have demonstrated their superb ability and have

received much attention for chemical process modeling. These techniques, for which

developing detailed knowledge of the process is of less concern, may overcome the

drawbacks of the mechanistic approach when dealing with complex and nonlinear systems.

Using AI-based methods, inherently qualitative variables in chemical processes like catalyst

deactivation in reactors can also be considered in the model, while these types of variables

are not possible to implement in mechanistic models.

3
3) Types Of Artificial Intelligence:

 Artificial neural network (ANN)

A neural network can be defined as a massively parallel-distributed processor made

up of simple processing units, which has a natural propensity for storing experiential

knowledge and making it available for use. It resembles the brain in two respects [3]:

(1) knowledge is acquired by the network from its environment through a learning

process; and (2) interneuron connection strengths, known as synaptic weights, are

used to store the acquired knowledge.

Figure 1.neural network with one hidden layer.

 Fuzzy logic (FL)

applying a fuzzy inference system. Fuzzy logic is an approach to computing based on

"degrees of truth" rather than the usual "true or false" (1 or 0) Boolean logic on which

the modern computer is based.

Fuzzy logic seems closer to the way our brains work. We aggregate data and form a

number of partial truths which we aggregate further into higher truths which in turn,

when certain thresholds are exceeded, cause certain further results such as motor

reaction. A similar kind of process is used in neural networks, expert systems and
4
 other artificial intelligence applications. Fuzzy logic is essential to the development of

human-like capabilities for AI, sometimes referred to as artificial general intelligence:

the representation of generalized human cognitive abilities in software so that, faced

with an unfamiliar task, the AI system could find a solution.[5]

 Genetic algorithm
Both the constrained and unconstrained optimization problems can be solved by GA,

which is a method based on a natural selection process that mimics biological evo-

lution [6]. In GA, a population of individual solutions is repeatedly modified by the

algorithm. GA is a part of the larger class of EA, which generates solutions for

optimization problems using techniques inspired by natural evolution, such as

inheritance, mutation, selection, and crossover. GA is an iterative procedure that

maintains a population of chromosomes representing different possible solutions to an

optimization problem. Each individual iteration is called a generation and in each

generation, the fitness of each chromosome is evaluated by a suitable fitness function.

By this approach, the optimal solution can be obtained in various fields of apparel

manufacturing such as apparel design, marker planning, PPC, production scheduling,

fabric and other materials location, and replenishment decision making in apparel

supply chain.

 Multi-Layer Perceptron (MLP).

comprises of three different layers: an input layer, one or more hidden layer, and an

output layer [7]. The number of neurons in the input and output layers is defined by

the number of input and output variables, respectively, while the number of neurons in

the hidden layer(s) is usually determined by trial-and-error.

5
4) Data processing and model development:

There are three ways in which the ANN learning takes place:

Supervised learning (training set)

An input stimulus is applied to the network, which results in an output response. This is

compared with the desired target response and an error signal is generated. The learning

in back-propagation networks is supervised.

Unsupervised learning (validating set)

During training, the network receives different input excitations and arbitrarily

organizes the patterns into categories. When a stimulus is later applied, the network

indicates the class to which it belongs and an entirely new class of stimuli is generated.

The learning in radial basis function networks is unsupervised.

Reinforced learning (testing set)

In this case, the network indicates whether the output is matching with the target or not-

a pass or fail indication. In other words, the generated signal is binary. This kind of

learning is used in applications such as fault diagnosis.[8]

6
Figure 2 steps in ANN model development process [9]

7
5) Application of AI techniques in chemical engineering

 Artificial Neural Network for Prediction of Chemical Oxygen Demand

Chemical Oxygen Demand (COD) is widely used as an important quantity for determining

the relative content of organic water pollutants in environmental monitoring and

environmental impact assessments. In view of this, the ability to predict the flux of COD

becomes important and relevant for monitoring and management of biological treatment

campaigns.

For experiment 8 input variables were considered:

 temperature,
 dissolved oxygen (DO),
 COD,
 total nitrogen (TN),
 total phosphorus (TP)
 suspended
 sediment (SS),
 transparency, and
 ammonia nitrogen (NH3-N)..

All this variables were divided in 3 sets of input

 the training set - in this set algorithm will understand the nature of each variable with

COD. 90% of input goes under training set.

 The validation set - used to decide when to stop training in order to avoid over-fitting

and/or which network structure is optimal;

 The Test Set - is used to assess the generalization ability of the trained model

8
 Performance of ANN was quantified by root mean square error, coefficient of correlation

and mean square error.

Result:

Figure 3 Series of the observed and predicted COD values during training phase.

These results suggest that COD can be accurately predicted using the seven observed

variables using the ANN method. Results show that DO, TP and transparency have the least

impacts on the prediction of COD and can be removed from the ANN model to simplify the

structure of the model, reduce the number of input variables, and reserve the model’s

capability to make good prediction. The method developed in this study can be easily applied

9
for forecasting other water-quality variables and provide a good tool to predict COD using

other variables that are easier to be measured. It is valuable to introducing this modelling

method into the process of river restoration for improving the effectiveness and efficiency of

the river restoration.[10]

10
References

[1] Moon, K., Ngai, E.,. The adoption of RFID in fashion retailing: a business value-added
framework. Industrial Management and Data Systems 108 (5), 2008, 596–612.
[2] Tayyebi S, Alishiri M. The control of MSF desalination plants based on inverse model
control by neural network. Desalination. 2014
[3] Majumder M Artificial Neural Network. In: Impact of Urbanization on Water Shortage in
Face of Climatic Aberrations, Majumder M (Ed). Springer Singapore, Singapore(2015)
[4] Araromi DO, Sonibare JA, Emuoyibofarhe JO. Fuzzy identification of reactive distillation
for acetic acid recovery from waste water. Journal of Environmental Chemical
Engineering. 2014;2:1394-1403
[5] Margaret Rouse. https://searchenterpriseai.techtarget.com/definition/fuzzy-logic
[6] Nayak, R., et al.,. Artificial intelligence: technology and application in apparel manu-
facturing. In: TBIS-APCC 2016. Binary Information Press, Textile Bioengineering and
Informatics Society. 2016
[7] Majumder M Artificial Neural Network. In: Impact of Urbanization on Water Shortage in
Face of Climatic Aberrations, Majumder M (Ed). Springer Singapore, Singapore, (2015)
pp 49–54.
[8] Fausett LV Fundamentals of neural networks: Architectures, Algorithms, and
Applications. Prentice-Hall, Inc., Upper Saddle River, NJ, USA. (1994)
[9] Holger R. Maier , Ashu Jain, Graeme C. Dandy , K.P. Sudheer Methods used for the
development of neural networks for the prediction of water resource variables in river
systems: Current status and future directions
[10] Gebdang B. Ruben & Ke Zhang & Hongjun Bao & Xirong Ma Application and
Sensitivity Analysis of Artificial Neural Network for Prediction of Chemical Oxygen
Demand, Water Resour Manage

11