Railway switching networks

A railway switching network is designed to control railway traffic control devices including railway switches and signals within a given area from a single point. Such control is exercised through the use of various track circuits which detect the presence of trains on a rail as well as other signal transmission devices.
History
The history of railway signalling and switching dates back to 1841: both railway semaphore signals as well as the methods behind railway routing and blocking developed in parallel, bringing into existence the railway switching network.
In 1856, the first mechanical centralization system came into existence in England. Such systems would later come to use electromechanical, electropneumatic and electrohydraulic methods of rail traffic control. Later, in Russia, the first mechanical centralization systems were built in a few stations on the Moscow - St. Petersburg line. In 1884, a mechanical centralization with rigid string action was constructed in the Sablino station.
Two electric and electromechanical centralization systems which appeared in the 1890s, Taylor's system in 1891 and Siemens system in 1893, were widely used. The first electromechanical centralization systems in Russia were built at the Vitebsk station of the Rigo-Orlovskaya Railway line in 1909 and at the Saint Petersburg station of the Moscow-Vindavo-Rybinsk railway line in 1914.
In 1934, a completely electrical system was installed in Gudermes station.
The next step for rail centralization was the development of semiconductor-based systems, a principle which was researched during the 1960s and 1970s with some prototypes being introduced in Germany and Japan. The first stations to implement the contactless centralization systems were Rezekne station of the Baltic Railway in 1968 and Obkhovo station of the October Railway in 1969.
The appearance of microprocessor based systems spurred the development of new stationary systems. The first system of a computer-based centralization system was built at Göteborg station in Sweden in 1975. The system, model number JZH-850, was manufactured by Telefon AB L M Ericsson in Mölndal and was based on a dual system with a realtime computer designed by Ericsson. One computer was live and the other a hot standby. This computer was produced until the year 2000.
The 1980s and 1990s marked the development and implementation of microprocessor-based systems in railway switching centralization.
Ericsson, SEL, SEG, Alcatel, JNR, Siemens, and DSI among others were the most active in the development of microprocessor-based centralization systems during this time period.
Needs and requirements of station trackage centralization
The speed of train processing at the stations plays a major role in railway permitting. Railway safety relies heavily on station operation. Switch and Signal Centralization is a kernel of stationary automation. Centralization provides logical interaction and blockage in accordance to safety standards, as well as economical and beneficial switch and signal control within a distance.
Today, railway centralization is implemented by means of Electromechanical circuits, sometimes with the help of electronics and microprocessors. The key element of any system is reliability. The system may fail 'safely' or 'unsafely'. Safe failure, is a failure which can result in a more prohibitive state of a centralization system (i.e. Red light instead of Green). Unsafe failure is when a systems fails giving more permissive state then should be (Green or Yellow instead of Red). Anything that may lead to unsafe failure, which can threaten thousands of lives, should be avoided (even though preventing from more advanced level). Failures must be safe, whenever they occur. Safe failure may cause delays, but no threat at all.
Anything that can lead to a failure of a system, (i.e. bad contact, operator's malpractice), should never cause an unsafe failure. That's why all the equipment ranging from relays and switches, should rather fail (not work) than to work wrongly.
Routing through station
* When station entrance trackage allows multiple routes from one point to another, basic route is a shortest path. It has the minimal number of crossing with other tracks, and allow the maximal speed.
* Any other route is called a variant route.
* Acceptance route provides train moving from the stretch to the station tracks.
* Transfer Routes are for trains which move from one station track to a consecutive.
* Departure Routes allow trains to go from a station track to the consecutive stretch
* Shunting Routes move the rolling units which are not ready to work in trains
Technical Characteristics of Centralization
First there were mechanical systems, where the signal and switch transition were done mechanically, using the traction string laid under station from the control point to the switches and signals. Signal and switch control was done by hand, and blockage was done using special portable keys (key dependency)
In mechanical systems, the human driven levers pulled wither rigid and flexible cords. Those cords were bound to switching mechanisms of switches and signals. Blocking dependencies were done by handles on the axles and rules with contacts, placed in dependency boxes which were the "brain" of the system.
The control can be made within 200-500 meters with rigid strings, and within 800 meters with flexible strings.
Later on electrohydraulic and electropneumatic systems were developed. The distance of the control of these systems is 1000-1200 meters. The executive devices such as semaphores, switches were electrically driven, the signal were transmitted by wires, however the switching was done still in mechanical dependency boxes as "brains". Not only such executive devices as Switches and signals were electrical, but also the decisions became to be done done completely by electromagnetic relays. Simply speaking, the "system acquired fully electric brains"
Such systems can control the switch network even with 5-6km distance. While, when this system uses code relays - the controllable distance is unlimited.
The even further generations of such systems are semiconductor and micro-processor-based. They are now being introduced in some places.
Electrical centralization systems, projecting basics
There are a few kinds of system of electrical centralization. They are different in complexity level, constructional design. It is defined by number of switches, signals, traffic loads, etc. When a low train load, no shunting work on a station is expected - it is needed to cheapen an electric centralization system. More perfect systems must be in use on stations with heavy train loads, with shunting works, etc.
The first electric centralization system, which was developed in 1936 featured local dependency and local power. All the equipment, which realized the dependency among switches, routes, and signals was placed in relay booths located by station inway' while an operating (dispatcher's) console was in station building.
In a central dependency system, the tools which perform route setting, closing and opening (excluding the contradicting route exception and other dependencies) are located in the station building. As a rule, all the relay equipment is inside a special relay room. All the modern systems are projected and developed with central dependencies.
Electric Centralization system with central dependency, yet local power was the only one, used on intermediate stations till the 1970s. It featured battery self-power for station lights, switch leads, and rail circuits. The "execution devices" of switches, lights, and rail circuits are in relay compartments, while all the "brains", i.e. dependency devices, are in station buildings.
This systems has some faults among accumulators, and the devices of external installation. That is why this system is in use on low-active branches with poor electric supply.
Controlling the system: The electric centralization prefers to use "direct wire" remote control. It features an individual signal line (i.e. wire) from the control point, which is dispatcher's desk, to a device.
Telemechanical coding is in use for distant districts of a station. It used coded centralization for route setting and, object condition acquisition. Station channels for this system require four-wire linear circuit.
Remote control is divided into individual control, and route control. Individual control features a separate button for every switch or signal light, while a route control features an automatic transition for all devices on the route - just by one key.
Closing the Rail Circuits methods divide the electric centralization systems on group contact closing, and sectional contact closing. In group closing systems, all sections open after the entire route was used (train passed.) In sectional closing, separate sections open while they become out of use.
Physically, electric centralization systems can be broken on "static" and "block". In block systems components are interconnected by detachable electric plugs, while in static systems, everything is connected by soldering (soft welding) of electric contacts. In block systems, separate dedicated circuit blocks are used for such operations as signal control, switch control, train sensors, etc.
Computerized and Semiconductor centralization
The major fault of the centralization system, dated back to 1960s was the usage of non-perspective base. The new perspective base appeared in 1970s when microprocessors started to be produced serially. A microprocessor is capable to do many things in data processing. It became universal and affordable. Let us take a look, how a microprocessor can be useful in solving Railway Centralization problems.
Since the train operations at the station are done simultaneously, the computer must work, emulating the parallel processes. The parallel process realization can be serial, conveyor, matrix, and multiprocessor.
In serial - the system has just one processor which process the processes in series. It is possible if the speed of the processing is much faster than that one of condition change of the station tracks. It creates the illusion of parallel calculations.
Functional processing features a couple of independent devices, which perform independent functions simultaneously.
Conveyor processing encounter that the process will be broken on a few steps, and then done either in parallel, or serially.
In matrix systems, there is a matrix of processors with a common control unit. The multiprocessor approach is done by a number of processors.

Main Menu

See Also

Main Menu