Loading...

Fault Recognition in a Four Stroke Internal Combustion (IC) Engine. An Artificial Neural Network (ANN) Based Approach

by Dr. Shankar Dandare (Author) Mayur R. Parate (Author)

Research Paper (postgraduate) 2015 91 Pages

Engineering - Automotive Engineering

Excerpt

CONTENTS

1. Significance of Fault Recognition System
1.1 Introduction
1.2 India’s Vehicle Growth
1.3 The Significance of Vehicle Maintenance

2. Literature Survey & Objective of Research Problem
2.1 Details of Literature Review
2.2 Observations from Related Research Work
2.3 Broad Objectives of the Proposed Systems

3. Data Acquisition
3.1 Experimental Setup
3.2 Collection of knowledge base
3.3 Mathematical Representation Parameters

4. Engine Faults and ANN Classifiers
4.0 Engine Faults Under Consideration
4.1 Spark Plug Fault
4.2 Piston Fault
4.3 Air Filter Fault
4.4 Working of Four Stroke Engine
4.5 Two-stroke and Four-stroke engines Comparison
4.6 Need of MATLAB and Simulink
4.7 Artificial Neural Network
4.8 Brief Introduction of ANN Based Classifiers

5. Design of Optimal Classifier
5.0 ANN based Optimal Classifier
5.1 Processing Using MATLAB
5.2 Classification of fault using SVM Classifier
5.3 Classification of fault using Two Layer FFNN

6. Experimentation using Simulink
6.0 Simulink
6.1 Signal Recording System using Simulink
6.2 Signal Conditioning
6.3 Parameter Estimation of Unknown Signal
6.4 Scatter Plots
6.5 Significance of ANN Based Classifier
6.6 Neural Network Inputs
6.7 ANN functions
6.8 Training ANN
6.9 Fault Recognition

7. Result, Conclusion and Future Scope
7.0 Testing of Spark Plug Fault Signal
7.1 Testing of Air Filter Fault Signal
7.2 Testing of Piston Fault Signal
7.3 Testing of Non Faulty Signal
7.4 Testing of Undetermined Signal
7.5 Conclusion
7.6 Future Scope

List of Figures

References

ABOUT AUTHORS

Abbildung in dieser Leseprobe nicht enthalten

Dr. Shankar Dandare working as a professor in the Department of Electronics and Telecommunication Engineering, Babasaheb Naik College of Engineering Pusad, India, since 1985, He received his M.Sc. degree in Applied Electronics in 1985, M.E. degree in Advanced Electronics in 1994 and Ph. D. degree in Electronics Engineering in 2015 from S. G. B. Amravati University, Amravati. India, He is a life member of many professional bodies like ISTE, AMIE and IETE. He has a total teaching experience of 33 years at UG and PG level of engineering. He is a reviewer of many international journals. He has published more than 25 research papers in reputed conferences and journals indexed in Scopus. He has author a book entitle “Fault Detection in Automobile Engines: A Soft Computing Approach” ISBN No: 978-3659-92847-5. His area of interest is Signal Processing, Optimization, Neural Networks and Soft Computing.

Mayur Rajaram Parate received his B.E. degree in Electronics and Tele-communication Engineering and M.E. degree in Digital Electronics from SGBAU, Amravati, India in 2009 and 2013 respectively. He is currently pursuing his Ph.D. degree from Visvesvaraya National Institute of Technology, Nagpur, India. His areas of interest are Signal Processing, Image processing and Computer vision.

Abbildung in dieser Leseprobe nicht enthalten

Preface

This book is very much useful for research scholar in the field of signal processing and neural network. The methods for fault recognitions in a four stroke petrol automobile engine have been explained very systematic manner. A very simple technique has been suggested to detect the multiple faults in an automobile. A soft computing technique with single sensor system has been tested for Spark plug fault, Air Filter Fault and Piston Fault. But the same type techniques can be extended to other types of two stroke and four stroke petrol engines, diesel engine, gas engine and electric engine also. Significance of this system is that it is less expensive. If this type of system is fitted in newly manufactured vehicle then it is possible to check the fault status of vehicle before induction in the market.

Chapter 1 of this book provides the details of growth of vehicular technology. As automobiles carry an important place in everybody's life and there are many benefits and drawbacks of vehicles and which is explained in detail. Day by day the rate of vehicles and rate of accidents is also increasing that details are also explained in this chapter. At the end, the significance of vehicle maintenance has been explained.

Chapter 2 illustrates the literature published in the related areas has been reviewed from the last two decades. The observations of reported research work have been explained. During the initial phase of the literature review, various fault detection systems in different areas of applications have been studied and analyzed. Subsequently, the scope of the survey was restricted to the research and development work specific to fault recognition systems with a view to prevent the vehicle from damaging. At the last broad objectives of the proposed systems are discussed.

Chapter 3 describes the data acquisition system. The details of signal capturing, signal decomposition and feature extraction system is also explained in this chapter. The mathematical expressions of features is also explained at the end.

In chapter 4 details of engine faults considered for fault recognitions is explained. Working and characteristics of four stroke engine, details of digital signal processing, Artificial neutral network and brief introductions of ANN based classifiers are also explained.

Chapter 5 describes the design of optimal classifier. Detail processing using MATLAB and classification of faults by using SVM based classifier and two layers FFNN are also explain in this chapter.

Chapters 6 illustrate the implementation using Simulink. The parameter extraction of unknown signal, selection of ANN based classifier and Fault recognition is explained in details in this chapter. After the training of a Neural Network, it is used to diagnose the status of a four-stroke internal combustion engine as faulty or non-faulty by applying unknown signal captured from the engine.

Chapter 7 covers the result, conclusion and future scope. Procedure for fault recognition of unknown signal is explained in this chapter. Finally architecture for proposed fault recognition system is given and details are also explained. Similar way system can be extended for other possible faults in engine and hence, had wide feature scope. Furthermore, the system architecture is generic and it can be applicable to any IC engine.

Acknowledgement: First of all, I wish to express my special thanks to my God for giving me a peaceful life and given me a chance to write this textbook. I wish to thank my Ph.D. Guide Dr. Sanjay V. Dudul, who infused the research temperament in me. He remained a great source of inspiration throughout my life. I express my deepest sense of gratitude towards him.

It is my pleasure to thank my Management Janta Shikshan Prasarak Mandal’s and Hon. President Shri Jai S. Naik, for their encouragement and wholehearted support.

I would like to thank Principal Dr. H. B. Nanvala, Prof N. P. Jawarkar, Head of Electronics Engineering Department and Prof A. M. Wankhade, Head of Mechanical Engineering Department, Babasaheb Naik College of Engineering, Pusad for providing laboratory facilities and necessary support as and when required.

I respectfully dedicated this work to my Parents who remained a key source of inspiration in my life.

Abstract

In recent times, research on effective Acoustic Emission (AE)-based methods for condition monitoring and fault recognition has attracted many researchers. They recognize that the advanced methods of supervision, fault recognition become increasingly important for many technical processes, for the improvement of reliability, safety and efficiency. The use of acoustic signals for fault diagnosis in four-strokes Internal Combustion Engine has grown significantly due to advances in the progress of digital signal processing algorithms and implementation techniques. The classical approaches are limited to checking of some measurable output variables and does not provide a deeper insight and usually do not allow a fault diagnosis. Engine problems are caused primarily by improper maintenance or fatigue caused by normal wear and tear and also worn out or clogged vehicle parts. The main cause of overheating of the engine, engine surging and other problems is noticed as worn out parts. The faults in Internal Combustion (IC) engine, reduces the performance, fuel average, smoothness also a change in engine sound is observed.

The faults in IC engines can be recognized and repaired based on engine sound and experience. However, as the engines are becoming more and more complex, getting expertise in fault recognition and localization is difficult, so there is a need of assistance system for fault recognition in IC engine, which will tell you about the possible fault based on the data provided to it.

The proposed system is model-based approach based on acoustic signals which is emitting from the engine. The Digital Signal Processing and Artificial Neural Network have been designed and implemented for fault diagnosis. Even though the complex properties of acoustic signals, effective features for fault detection can be easily extracted from the raw acoustic signals. This parametric method is based on parameter estimation, wherein a set of parameters is used to check the status of an engine and modal approaches are developed to generate several symptoms indicating difference between faulty and non-faulty status of an IC engine.

It is worthwhile to notice that the proposed system is an incipient fault detection and recognition system and must be attached to a newly produced engine, so that the fault can be detected at an incipient level. It is also suggested that the proposed Fault Recognition system can be extended to detect the maximum number of faults. It can be employed as a tool for quality control in production process to examine that the newly produced engine is fault free or not. It offers a guarantee to the customers that they are purchasing fault free vehicles. If the fault is developed afterwards, it can be easily detected, recognized and displayed on the front panel (Driver Information System) of the vehicle, so that it can be repaired at any desired service station. Therefore, the proposed Fault Recognition system will act as assistive tool for maintaining the vehicle in good condition that will save our time and avoid inconvenience. The proposed system is designed and implemented for four stroke petrol engine and experimentation is carried out to prove its validity as a fault recognition system.

Chapter 1 Significance of Fault Recognition System

1.1 Introduction

Vehicular technology has improved our lives and making it easier and comfortable. Automobiles carry an important place in everybody's life. There are many benefits and drawbacks of vehicles. The top most benefit is that it has reduced the travel time and one can reach to his destination in a blink of an eye. The drawbacks include its smoke creating air pollution and horn creating noise pollution. In future, the car technology would get more advanced and may produce flying cars that would be more useful for the people. These flying cars may overcome the technology of the airplanes and make the life of common people easier and comfortable.

Looking into the advancement of vehicular technology, day by day, the maintenance of the vehicle is becoming increasingly difficult because of the scarcity of skilled mechanic all over the world. Automobile engine is a complex system and sometimes the problems can be a bit tricky to diagnose. To diagnose the problem correctly, lots of knowledge and experience is needed. Engine problems are caused primarily by improper maintenance or fatigue caused by normal wear and tear and also worn out or clogged vehicle parts. Worn out parts may cause overheating of the engine, engine surging and other problems. When the problem arises and if it is not properly diagnosed and repaired in time then it may create some other severe problems and ultimately the engine may come to a halt. Therefore, automatic fault detection is very much essential to know the types of incipient faults being developed in the vehicles in advance.

1.2 India’s Vehicle Growth 1

MUMBAI: India's passenger vehicle and two-wheeler market is steering towards a strong double-digit growth as the economic environment improves amid a strong reform push by the Narendra Modi-led government, Munich-based global consultancy Roland Berger Strategy Consultants has said in a recent report.

According to the report, the Indian passenger vehicle market is expected to grow at a compound annual growth rate Compound Annual Growth Rate (CAGR) of 12% to 5 million units by 2020 as presented in Fig1.2 A. The two-wheeler market is also expected to grow at the same pace to 29.5 million units, while the commercial vehicle market will grow at a CAGR of 7% to 1.175 million units. All this would drive growth in the auto component sector to 14% CAGR, making it a $100-billion industry by the end of this decade, the report said.

Mr. Wilfried Aulbur, managing partner at Roland Berger Strategy Consultants, said key factors behind this growth are Gross Domestic Product (GDP) growth, growing disposable income and High Net Worth Individual (HNI) population. Mr. Aulbur said that the situation of 140 countries, it is found that a very strong co-relation between vehicle penetration and GDP per capita. His assumption is that the GDP growth will be at a CAGR of 7% till 2020 in India and if we grow at 7%, then the segment growth forecasted will be fairly evident."

Car penetration in India will grow at the fastest clip in the coming 5-7 years among the emerging markets, even outpacing China at 12% albeit on a low base. With the growing population and increasing wealth in the BRIC (Brazil, Russia, India and China) markets, the overall motorization rate will significantly increase until 2025.[1]

Abbildung in dieser Leseprobe nicht enthalten

Fig.1.2 A: Global Motorization Rates for year 2010 to 2025[2]

Skolkovo Institute for Emerging Market Studies, the report said that India is catching up in the future motorization rates. It is said that the penetration of motor vehicles per 1,000 inhabitants in BRIC countries will be faster than the 1%, 1.1% and 1.4% CAGR in developed markets of the US, the UK and Japan/Korea, respectively.

Abbildung in dieser Leseprobe nicht enthalten

Fig.1.2 B: Commercial Vehicle Forecast 2

For the commercial vehicle CAGR Forecast is depicted in Fig 1.2B in which, Small Commercial Vehicles (SCV), Light Commercial Vehicles (LCV), Medium Commercial Vehicles (MCV) and Heavy Commercial Vehicles (HCV) is shown from the year 2011 to year 2020.

By 2025, vehicle penetration in India is going to grow the fastest at 12.5% to 72 vehicles per 1,000 from 12 vehicles per 1,000 in 2010. China will grow at 10.1% to 187 vehicles per 1,000 inhabitants, Brazil at 3.8% to 221 per 1,000 and Russia at 4% CAGR to 388 vehicles per 1,000 inhabitants. Mr. Aulbur said that the GDP growth would be driven by good government policies and an action plan to implement them.

"There is a focus on details and there is a significant focus on performance within the administration, which is encouraging. The steps have been taken in the right direction. I would definitely like to see more action being taken to create the confidence. There are solutions in sight, which could have a major impact in the long term.

Abbildung in dieser Leseprobe nicht enthalten

Fig.1.2 C: Two Wheelers Growth in India 2

Abbildung in dieser Leseprobe nicht enthalten

Fig.1.2 D. Growth in Passenger Vehicles[3]

Fuel price deregulation, pushing for Goods and Services Tax ( GST), etc can have a big say in creating strong fundamentals for growth," Aulbur added.

According to Roland Berger, within the passenger vehicle space; micro cars will grow by 29%, mini cars by 6%, compact cars by 6%, super compact cars by 10%, mid-size cars by 15%, executive cars by 22%, premium and luxury cars by 84%, UV segment by 19% and vans by 17%. Within the commercial vehicle space; the heavy commercial vehicle segment will grow at the fastest pace at 12%, led by improving infrastructure, market trend moving towards higher tonnage vehicles and emergence of hub and spoke model in the country. The expected growth of passenger vehicles is shown in Fig 1.2D for financial year 2013 to financial year 2020.

The small commercial vehicle (SCV) space will grow at 5% on a high base. The goods segment will grow at 12%, whereas the van segment is likely to see a decline within SCV space with people preferring Multi-Purpose Vehicles (MPVs) and SUVs over vans. SUV: SUV is expanded as Sports Utility Vehicles and are big cars, which were designed to go over

Abbildung in dieser Leseprobe nicht enthalten

Fig.1.3 A: Road traffic death in India 1970 through 2014[4]

rough surfaces or off road. These can accommodate about 5-7 people. While sales in the commercial vehicle industry fell 20% in 2013-14, that in the passenger vehicle industry fell 6%. Capacity utilization in the commercial vehicle industry was 40% in that year, while in the passenger vehicle industry it was 60%. It was the two-wheeler segment that had the best capacity utilization at 74%, stemming the industry decline by growing over 14% to 14.8 million units.

The key drivers of the two-wheeler growth in India were favorable demographics, lower penetration, rising per capita income and poor public transportation.

The scooter segment will drive the two-wheeler growth on the back of shifting consumer preferences, the report said. The scooter segment is forecasts to lead the growth with 19.5% CAGR to 10.5 million units, followed by 9.4% growth in motorcycles that will translate to 17.9 million units. Mopeds will show a growth of 7% to 1.1 million units.

1.3 The Significance of Vehicle Maintenance National Road Traffic Injury Fatality

The World Report on Road Traffic Injury Prevention released by the World Health Organization that the magnitude of road accidents and fatalities in India is alarming as presented in Fig.1.3 A.

According to official statistics 141,526 persons were killed and 477,731 injured in road traffic crashes in India in 2014 (NCRB, 2015). However, this is probably an underestimate, as not all injuries are reported to the police (Gururaj, G., 2006, Mohan, D. et al., 2009). The actual numbers of injuries requiring hospital visits may be 2,000,000-3,000,000 persons. The basis for these estimates is given in later section. The situation in India is worsening and road traffic injuries (RTI) have been increasing over the past twenty years (Fig.1.3 B). This may be partly due to the increase in number of vehicles on the road but mainly due to the absence of coordinated evidence-based policy to control the problem. These data show that the number of fatalities has continued to increase at about seven per cent a year over the past decade except over the last couple of years.

Table1.3 shows eight countries with much higher vehicle ownership rates than India but lower RTI fatality rates.

National Fatality[5]

Fig.1.3 B shows the official estimates for total number of RTI fatalities and fatalities per 100,000 persons in India from 1970 to 2013 (NCRB). The total number of deaths in 2014 was 12 times greater than in 1970 with an average annual compound growth rate (AACGR) of 6%, and the fatality rate in 2014 was 5.2 times greater than in 1970 with an AACGR of 3.9%.

Abbildung in dieser Leseprobe nicht enthalten

Fig.1.3 B : Total number of RTI fatalities and fatalities per 100,000 persons in India[6]

Table1.3: Personal Vehicle ownership and Fatality Rate6

Abbildung in dieser Leseprobe nicht enthalten

There have been a few periods when the growth in RTI fatalities has decreased briefly and for a small amount, but the causes for the same are not known. This may explain the reason why the rate of growth in fatalities slowed down in India in the late 1990s and in the period 2010-2014 as these were also periods of low economic growth.

There is no indication of a long-term trend indicating that the increase in fatalities is going to reduce significantly in the near future.

Thus, based on projections of future income growth, they predict that fatalities in India will continue to rise until 2042 before reaching a total of about 198,000 deaths and then begin to decline. Koornstra uses a cyclically modulated risk decay function model, which in a way incorporates the cyclically varying nature of a society’s concerns for safety, and predicts an earlier date of 2030 for the start of decline in RTI fatalities in India. If we assume the average growth rate of 6% per year declines to nil by 2030, then we can expect about 200,000 fatalities in 2030 before we see a reduction in fatalities.

It is observed from the above Fig 1.3B that day-by-day fatalities are increasing. There are many reasons for this but vehicle condition is one of the reasons. Therefore, if the vehicle is maintained in good condition then it will avoid inconvenience, pollution and accidents too.

Chapter 2 Literature Survey & Objective of Research Problem

2.1 Details of Literature Review

Many researchers have given contributions towards fault recognition systems in IC engines. In most of the systems, use of multiple sensors is suggested to collect the data from IC engine to identify its faults. The sensors used are; air-mass flow sensor, intake-manifold pressure sensor, manifold temperature sensor, speed sensor, fuel actuator, etc. Furthermore, various processing and learning methods; Numerical Methods, Artificial Intelligence, Expert Systems, On-line Expert Systems and Neural Networks have been used to process and learn the data from sensors. The researchers’ contributions towards fault recognition systems are discussed below.

The literature published in the related areas has been reviewed for the last two decades and we have identified the strengths and weaknesses of the reported approaches, techniques and systems. During the initial phase of the literature review, various fault detection systems in different areas of applications have been studied and analyzed. Subsequently, the scope of the survey was restricted to the research and development work specific to fault recognition systems with a view to prevent the vehicle from damaging. The following subsections discuss the fault recognition systems in general and the research relevant to the prevention of vehicle maintenance, in particular.

In the year of 2008

S. K. Lee et al. demonstrated that the impulsive sound and vibration signals in machinery are often caused by the impacting of components and are commonly associated with faults. It has long been recognized that these signals can be gainfully used for fault detection. However it tends to be difficult to make objective measurements of impulsive signals because of the high levels of background noise. This paper presents an enhancement scheme to aid the measurement and characterization of such impulsive sounds called a two stage Adaptive Line Enhancer "ALE” which exploits two adaptive filter structures in series. The resulting enhanced signals are analyzed in the time-frequency domain to obtain simultaneous spectral and temporal information. In order to apply the two-stage ALE successfully the filter parameters and adaptive algorithms should be chosen carefully. Conditions for the choice of these parameters are presented and suggestions are made for suitable adaptive algorithms

R. J. Howlett et al. proposed a neural-network technique is described for the determination of the air fuel ratio in the engine. This is based on the neural interpretation of the signature which is formed from the voltage waveform at the spark plug. The neural network learning process requires that the air-fuel ratio is maintained accurately at a constant value during training; in order to facilitate this, a control system executing a fuzzy-logic paradigm is used.

In the year of 2007

Zheng Xiaojun et al. demonstrated the knowledge based diagnosis system for automobile engines. The system is based on the Hierarchical Diagnosis Principle, According to this principle, a complex diagnostic task can be divided into several simple ones and then solved step-by-step. Both deep and shallow knowledge are used in the system, and organised in two different knowledge bases:

1. A static knowledge base, which uses frames to describe the structure, symptom and fault information of the system to be diagnosed,

2. A dynamic knowledge base, which uses production rules and, special functions to describe various dynamic information for diagnosing the locations and causes of a system fault. The system employs a hierarchical and modular architecture which has two levels: a meta-level and an object-level. The knowledge base of the object-level system, according to the fault types and structure hierarchy of the system to be diagnosed, is divided into several independent knowledge sources which are controlled by the meta-level system.

S. K. Yadav et al. suggested a signal analysis technique for internal combustion (IC) engine fault diagnosis based on the spectrogram and artificial neural network (ANN). Condition monitoring and fault diagnosis of IC engine through acoustic signal analysis is an established technique for detecting early stages of component degradation. The location dependent characteristic fault frequencies make it possible to detect the presence of a fault and to diagnose on what part of the engine the fault is. The difficulty of localized fault detection lies in the fact that the energy of the signature of a faulty engine is spread across a wide frequency band and hence can be easily buried by noise. To solve this problem, the spectrogram for an integrated time frequency pattern extraction of the engine vibration is proposed. The method offers the advantage of good localization of the acoustic signal energy in the time frequency domain. Statistical parameters like, kurtosis, shape factor, crest factor, mean, median, variance etc., are used for feature extraction in time-frequency domain, and artificial neural network (ANN) was employed to identify the faults in IC engine.

In the year of 2008

Ho-Wuk et al. demonstrated that the sound quality degrades because of fault in the engine The impulsive sound signals in automotive engine are useful tool for their fault diagnosis since the impulses often occur due to irregular impacting. However detection of these impulsive signals is hindered by high levels of background noise such as fundamental frequency and harmonics of rotating speed and broadband noise. Under circumstances when a triggered signal is unavailable. They have proposed a scheme, it called the Two-Stage Adaptive Line Enhancer. It is formed by two ALEs and each with different roles. This method does not require any reference signal but use a delayed version (phase shifted) of the original signal as a reference signal. The output from this algorithm spectralized by time-frequency analysis to yield information about the time when impulsive occurs and simultaneously the frequency content of time signal.

Yingping Huang et al. a Bayesian Belief Network based method has been developed for automotive electronic system diagnostics. In contrast to the traditional troubleshooting flow diagram, the method proposed possesses two distinctive advantages: (i) the method is able to guide diagnostics on a probabilistic basis; (ii) the method is capable of simultaneously diagnosing multi-DTCs in an optimised way. These two advantages make automotive electronic system diagnostics more effective and accurate.

In the Year of 2009

S. K. Yadav et al. proposed a prototype-based engine fault classification scheme employing the audio signature of engines. In this scheme, Fourier transform and correlation methods have been used. Notably, automated audio classification has immense significance in the present times, used in both audio-based content retrieval and audio indexing in multimedia industry. Likewise, it is also becoming increasingly important in automobile industries. It has been observed that, real world automobile engine audio data are contaminated with substantial noise and out fliers. Hence, it is challenging to categorize different fault types in different engines. Accordingly, the present paper discusses a methodology where a set of algorithms checks the state of an unknown engine as either healthy or faulty. Fault categorizing algorithm is based on its cross and autocorrelation coefficient values. Appropriately, in this study, the engine amplitude-frequency values of fast Fourier transform are calculated and subdivided into bands to calculate the correlation coefficient matrix. The correlation coefficient matrix for the unknown engine is then calculated and matched with this “prototype” engine matrix to categorize it into a single or multiple fault(s).

Jian-Da Wu et al. proposes a fault diagnosis system for internal combustion engines using Wavelet Packet Transform (WPT) and Artificial Neural Network (ANN) techniques. In fault diagnosis for mechanical systems, WPT is a well-known signal processing technique for fault detection and identification. The used signal processing algorithm is inspired from previous work of speech recognition. In the preprocessing of sound emission signals; WPT coefficients are used for evaluating their entropy and treated as the features to distinguish the fault conditions. Obviously, WPT can improve the continuous wavelet transform (CWT) used over a longer computing time and huge operand. It can also solve the frequency-band disagreement by discrete wavelet transform (DWT) only breaking up the approximation version. In the experimental work, the Haar wavelets are used as mother wavelets to build and perform the proposed WPT technique. In the classification, to verify the effect of the proposed generalized regression neural network (GRNN) in fault diagnosis, a conventional back-propagation network (BPN) is compared with a GRNN network. The experimental results showed the proposed system achieved an average classification accuracy of over 95% for various engine-working conditions.

In the year of 2010

C. Angeli et al. proposes the evolution of the expert systems paradigm for fault diagnosis in technical systems and processes. Fault diagnosis is becoming one of the largest domains where expert systems are find application from their early stages. The process of diagnosing faulty conditions varies widely across to different approaches to systems diagnosis. The application of decision-making knowledge based methods to fault detection allows an in-depth diagnosis by simulating the human reasoning activity. Most of the past applications have been rule based while the automation of the diagnostic process including real-time data and/or modeling techniques added a new dimension to diagnostic task by detecting and predicting faults on line. Combination of expert systems technology with other artificial intelligent methods or with specific classical numerical methods adds more effectiveness to the diagnostic task.

In the Year 2011

H.G.P. Kabiri et al. proposed an effective Acoustic Emission (AE)-based method for condition monitoring and fault detection. Due to the complex properties of acoustic signals, effective features for fault detection cannot be easily extracted from the raw acoustic signals. To solve this problem, Fast Fourier Transform (FFT) is utilized. This method depends on the variations in frequency to distinguish different operating conditions of a machine. In this study, the intension is to categorize the acoustic signals into healthy and faulty classes. Acoustic emission signals are recorded from four different automobile engines in both healthy and faulty conditions. Even if the investigated fault is in the ignition system, the engine might be suffered from other possible problems as well that may affect the acoustic signals from the engine. The energy of FFT coefficients for acoustic signals at different frequency bands are calculated as representative features. Dimension reduction is performed on the dataset using Principal Component Analysis (PCA) method. The classification accuracy is validated and reported using 10-fold cross validation in which 10 percent of data is randomly selected for training and 90 percent for testing. The classification results are reported to be more than 80%.

In addition to above literature survey some of the researches work also tabulated in table 2.1.

Table 2.1: Summary of Literature Review Since 1995

Abbildung in dieser Leseprobe nicht enthalten

2.2 Observations from Related Research Work

From the critical review of literature published till date in the area of design and implementation of the Fault Recognition system the following observations are noticed.

- The fault diagnosis is not standardized among vehicle manufacturers.
- There is an uncertainty in scheduling vehicle.
- A complete diagnosis may need special equipments and trained technician’s help, which could make the diagnosis very expensive.
- Fault Recognition is difficult because of hybrid control systems.
- Multiple sensors are required to detect the multiple faults and hence, the system may be complex.
- Sensitive Sensors with high accuracy and precision are required.
- Maintenance and calibration of the sensors is difficult.

[...]


[1] http://economictimes.indiatimes.com/articleshow /44989399.

[2] Skolkovo Institute for Emerging Market Studies.

[3] Skolkovo Institute for Emerging Market Studies, the report

[4] NCRB-2015

[5] Road Safety In India,Status Report 2015, by Mr. Dinesh Mohan, Mr. Geetam Tiwari and Mr. Kavi Bhalla, IIT, Delhi.

[6] NCRB-2015

Details

Pages
91
Year
2015
ISBN (eBook)
9783668396753
ISBN (Book)
9783668396760
File size
3.1 MB
Language
English
Catalog Number
v346482
Grade
Tags
fault recognition four stroke internal combustion engine artificial neural network based approach

Authors

Share

Previous

Title: Fault Recognition in a Four Stroke Internal Combustion (IC) Engine. An Artificial Neural Network (ANN) Based Approach