Traffic Light :LOOP SENSOR Inductance Measurement in classification of vehicles

Thread Starter

kesarraghav

Joined Jan 8, 2016
9
Hi Everyone,
Hope you all are doing good!. I am working on Inductive loop sensors which are used to buried inside the ground in a form of saw cut loops and they are powered from a loop detector, when any vehicle passes over the loop then its inductance decreases and it gives a trigger to the loop detector. It works on the principle of mutual inductance as i said, the loop inside the ground is powered from loop detector(electronic unit), so it induces a magnetic field around it self, and when any vehicle passes its magnetic field changes and we get a trigger in our loop detector. I am using procon loop detector to perform this task. below link is the data sheet of the device.

http://www.proconel.com/Traffic,-Pa...es---Dual-channel-boxed-loop-detector-wi.aspx

I want to know:
1: What is the effect of temperature on the inductance of the loop and how much percent it affects the working of the loop detector.
2: How can we measure the loop inductance. Any tool?
3: Can I classify the vehicles using the loop sensor? means Trucks, cars, bikes

It would be great if you people gives your expert comments
 

#12

Joined Nov 30, 2010
18,224
This has been done a million times but you sound like you're designing a loop detector from scratch.
This usually indicates that your real concern has not been communicated.

1) Temperature has almost zero effect on inductance. 1b) The loop detector Product Brochure says it's good from -40C to +80C.
2) There are several ways to measure inductance, starting with an inductance meter, a signal generator and oscilloscope, and up through a network analyzer...in order of price.
3) That feature is not included in this model. Would you like to design one?
 

Thread Starter

kesarraghav

Joined Jan 8, 2016
9
No, I am not designing it from scratch! my task is to use this loop detector to classify the vehicles. In this loop detector , we gets one value Delta, which is small change in inductance and this value (Delta) varies for different vehicles. As per my analysis:

car: 3000-4000
Truck: 1500-2000
Bike: 50-200

But there are some false detection of 3 wheelers, These are the values which i get from loop size 2x1 metre and number of turns = 3.
Please suggest me , what is the significance of loop size in classification of vehicles,

1: Which loop size gives me the perfect classification.

http://www.gateinfo.org/knowledge/images/loop_1.jpg

2: AAC Fanatic! I want to know, how can i see the values in oscilloscope.
 

BReeves

Joined Nov 24, 2012
410
I have the understanding loop detection is old technology that is being phased out for many reasons. Traffic Lights and toll gates are now using cameras to detect size and types of vehicles.
 

blocco a spirale

Joined Jun 18, 2008
1,546
Road loops may be old technology but they are still one of the most reliable and maintenance-free forms of traffic detection. It may be possible to obtain more information from a camera but a camera requires regular maintenance which is expensive and can be difficult.

You can, of course, use more than one loop to help determine vehicle classification. I'm fairly certain that commercial devices use multiple loops that only detect the presence, or not, of a vehicle and classify it by vehicle length (rather than its electrical signature). Two adjacent loops provide enough information for vehicle speed and length to be calculated. Some vehicles may be incorrectly classified (e.g. a car towing a caravan may be classified as a lorry) but some margin of error is acceptable.

Road operators are often only interested in 4 or 5 categories of vehicle so there is no point over-complicating things.
 
Last edited:

Sensacell

Joined Jun 19, 2012
3,432
The problem is also compounded by the relative POSITION of the vehicle in the loop, if the car is half-way, it might look like a bike.
The output is one-dimensional while the problem you are solving has many dimensions.
 

BReeves

Joined Nov 24, 2012
410
Just going by what a friend experienced that use to have a business installing loops for traffic lights in Dallas. He was basically put out of business because they went to cameras and no longer needed his service. Also noticed all new traffic lights that go up around Tulsa are using cameras and they are slowly adding cameras to existing lights. Government entities are usually not the most frugal with money but even then installation cost and maintenance must be a factor in the move to cameras.
 

blocco a spirale

Joined Jun 18, 2008
1,546
I speak from experience, cameras seem like a good idea until they've been in use for a few months and the spiders and insects have made their homes in front of the lens and the traffic and road dirt has fogged the glass. They also perform poorly when the roads are wet and lights reflect off the surface of the road and cause multiple false detections. There is, perhaps, more cost involved installing a loop but once in it will operate for years without any maintenance.
 

Alec_t

Joined Sep 17, 2013
14,280
what is the significance of loop size in classification of vehicles
If the loop is very small then the sensor output is likely be the same for any vehicle, so a single-loop system would need a loop about as big as the largest vehicle of interest to aid classification. Multiple loops, as Blocco mentions, would be more useful for classification.
Which loop size gives me the perfect classification.
I doubt you can achieve perfection ;).
 

Kermit2

Joined Feb 5, 2010
4,162
The amount of iron varies widely in the same type and size vehicles. A Ford Mustang and a Toyota Corolla are about the same size shape and weight. Owing to the different alloys of steel and of aluminum used. These two vehicles might give very different signals when placed in the exact same position above your loop detector.
In short. what you want is not possible.
You cannot get there from here.
 

alfacliff

Joined Dec 13, 2013
2,458
the loops respont differently to viehicles of larger mass, or even height as different size viehicles. the ones we use here at work can be set for sensativity and also for continous detect or pulse output. if left on pulse output, they only give a pulse for each viehicle, and will not tell you if anything is settng on the loop. the sensativity can cause it to ignore small things like bicycles. usually there is a third control, to set the base frequency so neary loops will not interfere.
 

Thread Starter

kesarraghav

Joined Jan 8, 2016
9
Thanks all for your expert comments. I am researching on this topic and in my research, I have analysed that Inductance changes when we increase the number of turns in the loop sensor. Moreover, the loop size is directly proportional to the height of detection of the vehicle from the ground.
Height of detection = 2/3 ( shorter side of loop )
If loop size is 6x3 " feet then This loop will sense up to 2 feet.
I have designed several loops like rectangular, square, quadruple etc and varying the number of turns to find out the best result. But I am still in search of perfection.
The change in the inductance of the loop when vehicle passes over it depends on following factors which I have analysed:
1: Speed of vehicle
2: Position of the vehicle in the loop
3: size of loop and number of turns
4: Length of loop wire extended up to loop detector

I request you all to give your expert comments on my analysis . How can i improve the classification part.
 

blocco a spirale

Joined Jun 18, 2008
1,546
I just did, in post #6. Vehicle classifications, in most cases, can be determined by vehicle length.

It all depends on how many classifications are required, it should be possible to determine at least four e.g. motorcycle, car, goods vehicle, articulated goods vehicle.
 
Last edited:

AnalogKid

Joined Aug 1, 2013
10,987
Something not explicitly mentioned yet is that to accurately classify a vehicle based on nothing more than its magnetic signature, the vehicle must be stopped. There is no way a single sensor can measure the length of a moving object unless you have a very accurate indication of its speed. A slow Prius and a fast moving van look exactly the same to a single point sensor. Two sensors close together frequently are used to trigger speed trap cameras, and might be able to estimate the length of the vehicle if tehre is no reason for the vehicle to be speeding up or slowing down at that location. But it cannot tell you the width, height, or weight, things needed to classify a vehicle.

ak
 

Thread Starter

kesarraghav

Joined Jan 8, 2016
9
Thanks, Man,
AnalogKid
I completely agree with your point. When I stop the vehicle on the Loop at a perfect position, It gives a huge difference in the values of the car, Truck, and Bike. But when the same vehicles are in motion I got a slight change in values.
I can conclude my analysis with this decision: That either a vehicle should be stopped over the loop or there should be two loops which I have not been worked on.
blocco a spirale I will try with two loops now. :)
 
Top