Fault detected at the earlier stage at any part of the three phase induction motor may protect the machine from catastrophic failure. Due to that reason early stage detection of fault plays an important role in nondestructive preventive maintenance of three phase induction motor. So, early detection of fault could save the machine and as well as protect the system from the total shutdown. This health monitoring could be done by using various sensors also, but that system needs various sensors connected directly to the machine. So, the different sensors connected to the machine create the system complicated and sometime make the system inefficient to detect the fault at the earlier stage. Therefore, the researchers have taken their interest in the field of health monitoring for the preventive maintenance of the machine which are widely used in the industry by using contactless method. Particularly, in three phase induction motor, different type of bearing faults occur because of the mass unbalance on motor shaft. In such cases, a mechanical load distributed asymmetrically over the shaft, causing displacement of the center of mass of the elements coupled to the motor from the rotation center of the machine. This asymmetric distribution generates vibrations and strokes. This unexpected friction produces some of the power loss and that loss is coming out in the form of heat. Similarly, in all the faults in the induction motor produce some of the heat in the particular parts of the machine. Different types of faults in induction motors that could be electrical faults or mechanical faults. Electrical faults include different winding faults and rotor faults. Whereas the mechanical faults could be bearing faults and eccentricity faults. In all the above said faults the machine directly or indirectly will gradually produce the heat in the concerned part of the machine. So, thermal imaging is one of the pioneer methods to detect the fault at the earlier stage without having any contact with machine. Now a day, the infrared (IR) thermograph technology has been recognized and accepted as health monitoring method by researchers. Moreover, this method is one of the most popular gained more recognized and accepted due to its non-contact and nondestructive features of inspection. It is very quick and trustable monitoring system which can monitor the induction motor without any interference to the whole system. In IR thermograph based technique, health monitoring is performed by making the analysis of the thermal images captured by infrared camera. It is very well known fact that durability of any electrical equipment is notably reduced as temperature rises. Infrared (IR) thermograph technology offers many advantages over any conventional methods such as prompt response times, ample temperature ranges, highly reliable, harmless, high spatial resolution, and very lucrative approach for the health monitoring of electrical power systems machinery.
In this research work health monitoring of an induction motor by using infrared (IR) thermograph technology has been developed for both real time and off line application. Continuous monitoring of induction motor gives the real time information, which is useful for health monitoring of running induction motor. On the other side off line monitoring gives the information of standstill contacts which is useful for the health monitoring of other apparatus like fuse cabinet, inductive heating, corrosion, electrical panel and all those which is stand by connected to motor.
Due to the consequence of liberalization, the new investment in the electrical machinery has come down over past two decades. A lot of electrical machines specially three phase induction motors have been working well beyond their anticipated life and moreover, they are operating under rising stress. As these machines are often exposed to hostile environments during operation, this leads to deterioration and hence work beyond the specification. This leads the situation to work under unhealthy condition. As a corollary, new technologies must be exposed to allow electrical machine to better fit under such circumstances and also be economically acceptable and reliable. A good amount of attention is paid against the health monitoring of induction motor for the quality and uninterruptable drive system. The equipment on which whole drive system depends is induction motor and the monitoring of induction motor in efficient way is still a big task. In most of the existing health monitoring systems, monitoring of induction motor is done through the current signature analysis. But this technique is tedious and costly procedure to detect faults. Few mathematical models and fuzzy rule based techniques are also now a day’s popular to detect the health of the motor, but, these techniques also require some measured data to interpret and predict. So, the non invasive temperature monitoring of three phase induction motor and its components with less complexity and high reliability is a prime necessity with existing system to overcome the maintenance cost and revenue loss during the fault of the system. The infrared (IR) imaging technique makes non invasive type health monitoring systems for the induction motor which is more reliable for prediction of the health of electrical machine. This IR technique is also useful to monitor the health of any household electrical appliances, real time thermal monitoring of power system devices like transformer, measurement of excitation winding temperature in synchronous generator, and real time high temperature measurement in control industry. The design of visual inspection system (VIS) based on chargecoupled device (CCD) and complementary metal oxide semiconductor (CMOS) cameras are already into existence for the monitoring of different processes, objects, sorting, and quality check but designing VIS based on IR cameras for health monitoring system are very rare.
In most of the papers discussed above, only invasive or pointed temperature is considered and no non-invasive monitoring and controlling systems were proposed in any research. But, in our case, RGB color model is used to detect the faulty condition of the motor. By using the R-mean G-mean and B-mean we find the abnormal temperature rise in a particular part or area of the machine. Graphical analysis of the thermal image also is done for better understanding of the abnormal temperature rise of any part of the machine. Two IR imaging based visual monitoring and controlling systems are proposed: Non Invasive off Line Visual Inspection System (NIOLVIS) and Non Invasive Real Time Visual Monitoring System (NIRTVMS), for off line and real time applications respectively. The proposed systems have the ability to identify the area of the hotspots in the induction motor and make decisions accordingly to keep other part of the machine healthy. The proposed methods are simulated for a large number of thermogram images of induction motor and performance are analyzed using statistical and geometrical features. This method is executed by using MAT LAB.
A. Thermal Image
Infrared camera will not have the ability to measure the body temperature directly of the machine. The camera detectors are very much sensitive to the luminance which is emitted from the object whose photo is to be taken. This luminance is directly proportional to the temperature of that particular object. That kind of image captured by the infrared thermography camera is called thermal image. This camera is able to capture electromagnetic spectrum within infrared bands,0.78–1000 μm. A thermal image is a function of radiated energy on an inspected object. This thermal image can be converted into digital form and can be represented in matrix form for computational processing as follows:
B. Image Thresholding
Image thresholding is commonly used for the process of image segmentation. Thresholding is also very useful method for the detection of fault of any substance by IRT image.Thresholding is a process to separate objects from its background in a digital image. Histogram is the main tool in this separation process. Say that the grey level corresponds to an image f (x, y) that is composed of lights objects on a dark background, in such a way that object and background pixels have grey levels grouped into two dominant modes. Extracting the object from the background is performed by selecting a threshold T that separates these modes. A threshold image g(x, y) is defined as
C. RGB colour model
The RGB (red green and blue) is colour model based on the three basic colours red green and blue. These three colours are mixed together in different amount to produce a broad array of colours. The name of the model comes from the initials of the three additive primary colours red, green, and blue. For the representation and display of images in electronic systems, such as televisions and computers and also in digital cameras RGB colour model is used. In infrared thermal camera also this RGB model is used to display the thermal image. The first experiments with RGB in early colour photography were made in 1861 by Maxwell himself, and involved the process of combining three colour-filtered separate takes. In modern technological era we can analyse the digital image by various methods. A broad area for the researchers is now a digital image processing. We can very easily analyse the image taken by any infrared thermal imaging camera or any simple digital camera by using the MATLAB.We can plot the whole image in a numerical matrix form and easily can identify the each pixel. Each pixel contains three colours red, green and blue as the theory of RGB code. We can identify the intensity of colour of each pixel by the value of individual red green and blue (RGB). MATLAB can be used to calculate the average value of each individual red, green and blue (RGB). We can make any comparison between images. In Infrared thermal imaging camera the temperature is identified based on its colour. The image will be brighter for the higher temperature in the thermal imaging camera. By calculating the average value of red, green and blue (RGB) in a pixel in the thermal image we can detect the hotter place or the hottest place in the image. Even we can identify the hottest pin point and the location of that pixel where the hottest point situated. Here one thermal image of a running induction motor is shown in Figure 1 by using Thermal imaging Camera.
A. Test Bench
The experimentation is done in Electrical Engineering lab at Dayalbagh Educational Institute (Deemed University) Agra, India. The experiment setup is consisting of one induction motor having a healthy cooling system with the arrangement through which fault in the cooling system can be created. The specification of the motor are three phase Squirrel cage induction motor 230V, 50Hz, 3HP, inbuilt fan type cooling. A voltage –current-frequency (VIF) meter, one infrared thermal imaging camera with a proper fixed camera stand and one computer with MATLAB software installed. The real time image is acquired with the help of infrared thermal imaging camera for both healthy and faulty condition of the cooling system at different load conditions. For both the condition data are sent to computer. MATLAB imports these signals and graphs are drawn for all the parameters at faulty and healthy condition. Graphical analysis is also done for the final declaration of the health condition of the induction motor.Fig.2 shows the experimental setup of health monitoring system of induction motor using infrared thermal imaging camera.
B. Infrared Thermal Camera Specification
Used camera specification for the thermal imaging for experimental purpose of induction motor is shown as follow.
– Focus: Manual
– Field of view / min focus distance: 18° x 13° / 0.3m Thermal sensitivity: ≤ 0.1°C@30°C – spatial resolution (IFOV): 1.9 mrad Frame rate: 50/60 Hz
– Spectral range: 8 – 14μm
– Electronic zoom: 2X
– Array size / format: 160 X 120
– Detector type: Uncoiled FPA microbolometer.
– Temperature ranges : -20°C ~ + 350°C, optional up to +600°C or 1000°C
– LCD: Built-in-high resolution Color 2.5O LCD
– Setup functions: Date / time, temperature unit, language
– Emissivity correction: Variable from 0.01 to 1.0
– Ambient temperature correction: Automatic correction according to user input
– Atmospheric transmission correction: Automatic correction according to user input object distance, relative humidity, ambient temperature.
– Accuracy: ± 2°C or ± 2 % of reading, whichever is greater.
Laser pointer: Class 2, 1mw / 635nm (red)
Storage mode: Automatic / manual single image saving
File format – thermal: JPEG, 14 bit thermal image with measurement data
For the proposed method, we have a system i.e. one induction motor with the arrangement that we can create the fault in the cooling system of that motor easily. One digital camera is fixed on the stand to capture the infrared digital images. First we acquired the infrared thermal images of the motor from the healthy motor at different load conditions. Then we create the fault in the cooling system of motor manually.
Then we acquired the infrared thermal images the faulty motor at different load conditions. All the images are taken from a fixed location from the camera stand. Then we plot the images in the matrix form by Matlab and also crop all the images to a fixed size to consider the exact motor portion. We find the hottest part of the machine by finding the RGB average and maximum of the image pixels.RGB plots are plotted to visualize the results better. After making the comparison between the healthy and faulty part of the motor we can conclude the fault. For all the analysis only thermal images are used in MATLAB.
Result and Analysis
A. RGB average of Healthy motor image at different load
If the load is increased then the heat is increased with the load. Trends show that read mean and green mean both are also increased with the increase in heat, but the blue mean is decreased with the increase in heat.
B. At healthy condition
(a) No-load image and its RGB Analysis
(b) Quarter Load image and its RGB Analysis
(c) Half Load image and its RGB Analysis
C. At faulty condition (Cooling system failed)
(a) At no load (cropped) image and its RGB Analysis
(b) Quarter Load image and its RGB Analysis
(c) At half load (cropped) image and its RGB Analysis
The purpose of this paper is to diagnose the health condition of the induction motor by using an innovative technique where the whole motor will be monitored without having any physical contacts.
Here no sensor is used to get any physical parameters. So, the system is very simple and chances of failure are almost nil. It is very much important for an engineer to monitor the induction motor while it is working in a system, because any fault generated in induction motor can be the reason of excessive heating.
Excessive heat produced in the machine can burn the windings of the motor and the situation will be very harassing. In this method all the sensor which is normally used in conventional condition monitoring is replaced by an infrared digital camera alone. The system is almost maintenance free and analysis is very particular to the every part as if the whole motor condition is vivid in front of eyes.
The analysis results showed that the proposed method is able to monitor the health condition of the induction motor. The method described provides a promising way to establish potential metrics for the description of the health of an induction motor. Therefore, it is desirable to develop an on line health monitoring system for the induction motor based on the above method and realize on-line health evaluation of each part of induction motor. With such a function, the critical failure of induction motor systems can be avoided, and the reliability and efficiency of motor can be increased.