RAILWAY TRACK SECURITY SYSTEM

The main problem is about railway track crack detection at the earliest to avoid accidents. Therefore it is essential that such problems must be communicated immediately to the concerned authorities by using GSM technology for appropriate action.

This project uses a microcontroller from 8051 family. The primary objective of this project is to detect the crack in the railway track and alert the nearby station through effective and highly reliable communication mode. To demonstrate this project, two rails forming the part of the track are made using a pair of wire which is wired with a detachable jumper in between each wire/track. Removing the detachable jumper creates a fault in the respective track; otherwise it is generally shorted by the jumper wire to simulate healthy track condition.

Removing the jumpers result in driving transistors delivering a different logic to the controller. The program thereafter takes over to send an SMS through GSM modem interfaced through MAX232 level shifter IC to the microcontroller. An LCD is also interfaced with the MC to display the status of GSM and track condition. Thus the proposed model is designed to recognize the cracks in the railway tracks and provides instant information to the concerned railway authorities. The power supply consists of a step down transformer 230/12V, which steps down the voltage to 12V AC. This is converted to DC using a Bridge rectifier. The ripples are removed using a capacitive filter and it is then regulated to +5V using a voltage regulator 7805 which is required 

Convert 1 to 5V signal to 4- to 20-mA output

Despite the long-predicted demise of the 4- to 20-mA current loop, this analog interface is still the most common method of connecting current-loop sources to a sensing circuit. This interface requires the conversion of a voltage signal—typically, 1 to 5V—to a 4- to 20-mA output. Stringent accuracy requirements dictate the use of either expensive precision resistors or a trimming potentiometer to calibrate out the initial error of less precise devices to meet the design goals.

Neither technique is optimal in today’s surface-mounted, automatic-test-equipment-driven production environment. It’s difficult to get precise resistors in surface-mount packages, and trim

ming potentiometers require human intervention, a requirement that is incompatible with production goals.

The Linear Technology LT5400 quad matched resistor network helps to solve these issues in a simple circuit that requires no trim adjustments but achieves a total error of less than 0.2% (Figure 1). The circuit uses two amplifier stages to exploit the unique matching characteristics of the LT5400. The first stage applies a 1 to 5V output—typically, from a DAC—to the non - inverting input of op amp IC1A. This voltage sets the current through R1 to exactly VIN/R1 through FET Q2. The same current is pulled down through R2, so the voltage at the bottom of R2 is the 24V loop supply minus the input voltage.

This portion of the circuit has three main error sources: the matching of R1 and R2, IC1A’s offset voltage, and Q2’s leakage. The exact values of R1 and R2 are not critical, but they must exactly match each other. The LT5400A grade achieves this goal with ±0.01% error. The LT1490A has less-than-700-μV offset voltage over 0 to 70°C. This voltage contributes 0.07% error at an input voltage of 1V. The NDS7002A has a leakage current of 10 NA, although it is usually much less. This leakage current represents an error of 0.001%.

The second stage holds the voltage on R3 equal to the voltage on R2 by pulling current through Q1. Because the voltage across R2 equals the input voltage, the current through Q1 is exactly the input voltage divided by R3. By using a precision 250Ω current shunt for R3, the current accurately tracks the input voltage.

The error sources for the second stage are R3’s value, IC1B’s offset voltage, and Q1’s leakage current. Resistor R3 directly sets the output current, so its value is crucial to the precision of the circuit. This circuit takes advantage of the commonly used 250Ω current-loop-completion shunt resistor. The Riedon SF-2 part in the figure has 0.1% initial accuracy and low temperature drift. As in the first stage, offset voltage contributes no more than 0.07% error. Q1 has less than 100-nA leakage, yielding a maximum error of 0.0025%.

Total output error is better than 0.2% without any trimming. Current-sensing resistor R3 is the dominant source of error. If you use a higher-quality device, such as the Vishay PLT series, you can achieve an accuracy of 0.1%. Current-loop outputs are subject to considerable stresses in use. Diodes D1 and D2 from the output to the 24V loop supply and ground help protect Q1; R6 provides some isolation. You can achieve more isolation by increasing the value of R6, with the trade-off of some compliance voltage at the output.

If the maximum output-voltage requirement is less than 10V, you can increase R6’s value to 100Ω, affording even more isolation from output stress. If your design requires increased protection, you can fit a transient-voltage suppressor to the output with some loss of accuracy due to leakage current.

This design uses only two of the four matched resistors in the LT5400 package. You can use the other two for other circuit functions, such as a precision inverter, or another 4- to 20-mA converter. Alternatively, you can place the other resistors in parallel with R1 and R2. This approach lowers the resistor’s statistical error contribution by the square root of two.

USB Booster..

As you probably know, the USB 2.0 ports can deliver up to 500 mA that means about 2.5W. But sometimes you might need more power to connect an external HDD or other peripherals and the USB ports just cannot deliver enough current. In this case you can buy USB hubs that have an external power adapter required to boost the power or you can build a simple or complex circuit that can do the same thing.  We are providing a very simple design involves the use of the 7805 voltage regulator that can deliver 5V and 1A.

The USB serial bus can be configured for connecting several peripheral devices to a single PC. It is more complex than RS232, but faster and simpler for PC expansion. Since a PC can supply only a limited power to the external devices connected through its USB port, when too many devices are connected simultaneously, there is a possibility of power shortage. Therefore an external power source has to be added to power the external devices. In USB, two different types of connectors are used: type A and type B. The circuit presented here is an addon unit, designed to add more power to a USB supply line (type-A). When power signal from the PC (+5V) is received through socket A, LED1 glows, opto- diac IC1 conducts and TRIAC1 is triggered, resulting in availability of mains supply from the primary of transformer X1. Now transformer X1 delivers 12V at its secondary, which is rectified by a bridge rectifier comprising diodes D1 through D4 and filtered by capacitor C2. Regulator 7805 is used to stabilize the rectified DC. Capacitor C3 at the output of the regulator bypasses the ripples present in the rectified DC output. LED1 indicates the status of the USB power booster circuit. Assemble the circuit on a general purpose PCB and enclose in a suitable cabinet. Bring out the +5V, ground and data points in the type-A socket. Connect the data cables as assigned in the circuit and the USB power booster is ready t o function.

Bioinformatics Applications

Machine Learning Methods in Bioinformatics Applications

The proteomics is an important domain where machine learning techniques are applied in bioinformatics. In the proteomics, two main applications of computational methods are protein structure prediction and protein function prediction. Generally, the first is an optimization problem and the second is a classification problem. Evolutionary algorithm (EA) based methods are the main optimization technologies for protein structure prediction, such as genetic algorithm (GA), estimation distribution algorithm (EDA), etc. Supervised and unsupervised classification methods are often used to predict protein function, such as Clustering, SVM, NN, etc.

SVM

 Support Vector Machines for Real-world Pattern Recognition

SVM is a nonlinear pattern recognition algorithm based on kernel methods. In contrast to linear methods, kernel methods map the original parameter vectors into a higher (possibly infinite) dimensional feature space through a nonlinear kernel function. Without need to compute the nonlinear mapping explicitly, dot-products can be computed efficiently in higher dimensional space. The dominant feature which makes SVM very attractive is that classes which are nonlinearly separable in the original space can be linearly separated in the higher dimensional feature space. Thus SVM is capable to solve complex nonlinear pattern recognition problems. Important characteristics of SVM are its ability to solve pattern recognition problems by means of convex quadratic programming (QP), and also the sparseness resulting from this QP problem.

Yin-Yang EAs Balancing Adaptivity and Diversity

Evolutionary algorithms (EAs) are search and optimization algorithms based on the principles of natural evolution, which have found successful applications in bio genetics, computer science, engineering, economics, chemistry, manufacturing, mathematics, physics and other fields.

   In applying EAs to solve large-scale real world problems, however, confronted with the conflict between accuracy and speed, EAs often result in an unsatisfactory compromise. Furthermore, one of the commonest difficulties encountered is premature convergence.

Quasi-ARX Modeling and Identification

Neural networks (NNs) and neurofuzzy networks (NFs) have been proved to have universal approximation ability. They can learn any non linear mapping. Many non linear ARMAX models have been proposed based directly on NNs and NFs. However, system identification is always followed by certain applications such as system control and fault diagnosis. From a user's point of view, NNs and NFs are not user-friendly since they do not have structures favourable to the applications of system control and fault diagnosis. To solve this problem, it is natural to consider a modelling scheme to construct models consisting of two parts: macro-part and kernel-part. The macro-part is a user-friendly interface constructed using application specific knowledge and the nature of network structure; efforts in this part are made to introduce some properties favourable to certain applications, while to embed the resulted model complexity in the coefficients. The kernel-part is a flexible multi-input multi-output (MIMO) non linear model such as NN and NF, etc. which is used to represent the complicated coefficients of macro-parts. Non linear models constructed in this way are expected to be user-friendly and to have excellent presentation ability.