AU7228081A - Measurement of speed and/or length - Google Patents

Description

"MEASUREMENT OF SPEED AND/OR LENGTH"

TECHNICAL FIELD

The present invention relates to speed and length measurement;. It has particular application to the determination and -speed or length measurements of hot or cold steel product in a rolling mill but it is not limited to this particular application, and it could be applied to speed and/or length measurements of many other materials in solid, liquid or slurry form or, more generally, to the measurement of relative speeds between various bodies. ^* The measurement of product speed and length are of major importance in many production processes.- Specifically, in the steel rolling process, speed data is required in order to synchronise such items of equipment as flying shears, diverting switches, water sprays, coilers, loopers etc. , as well as a means of controlling the thickness reduction of each pass through, a mill. Product length data is also critical to the on-line cutting of continuous or semi-continuous product to meet constraints set by customer requirements, run-out table lengths and materials handling facilities. BACKGROUND^* ART

Conventional speed and length measuring- systems rely on,contacting measurement techniques, either by tachogenerator or other means of determining th.e rotational speed of the reduction rolls or of a specially installed idler roll which has been provided with some means of contacting, and hence being driven by, the product. These techniques suffer from errors from several sources including variations or uncertainty in roll diameter, particularly in profiled rolls, unknown slip factors between the roll and the product and human errors such as setting the wrong diameter data into the device which calculates speed^* for a wide range of roll sizes. The invention described here alleviates all of these problems by enabling use of a non-contact, direct reading instrument of fixed calibration. Direct reading refers to the speed data being determined directly from the linear motion of the product and not by a rotating interface. DISCLOSURE OF INVENTION

According to the invention, a method of determining a relative speed between two bodies generally comprises deriving from a pair of detectors fixed relative to one of the bodies first and second time sequences of signals characteristic of physical variations in the other body at two locations spaced apart in the direction of movement of said material, determining the time delay between the two sequences of signals for which the two sequences most

OMPI IPO closely match one. another, and dividing the distance^' between said locations by said time delay to determine, the speed of the material.

The two time sequences of signals derived from the detectors may initially be in the form of continuously fluctuating electrical signals. Said time delay may then be determined by the following steps: converting the two sequences of electrical signals to digital form; storing the two sequences of signals in digital form; making separate comparisons of each signal of the first sequence with digital signals of the second sequence derived at differing time lags behind said signal of the first sequence; and determining which of said differing time lags produces maximum correlation between the compared signals of the first and second sequences.

The making of the separate comparisons of each signal of the first sequence with signals of the second sequence may be carried out by comparing a first digital signal of the first sequence successively with digital signals of the second sequence derived at successively greater time lags behind said first digital signal; and comparing successively later signals of the first sequence each successively with digital signals of the second sequence derived at successively greater time lags behind said later signal of the first sequence.

Alternatively, said separate comparisons may be carried out by comparing a group of successive signals of the first sequence successively with" the groups of successive signals of the second sequence corresponding to successively greater time lags behind the group of the first sequence.

The invention further provides a method for determining a length measurement of a linearly moving material which comprises determining by the above defined method instantaneous material speed values throughout a time interval during which the material passes the detectors and integrating the instantaneous -speed values over said time interval. The material ay- be allowed to move through a particular known distance prior to the time interval for which the instantaneous speed values are integrated and the length measurement derived by addition of the known distance with the integrated speed values. The invention further provides apparatus for determining a relative speed between that apparatus and another body comprising a pair of detectors to derive first and second time sequences of signals characteristic of physical variations in said other body at two locations spaced apart in the direction of said movement and signal processing means for determining the time delay between the two sequences of signals for which the two sequences most, closely match one another and for dividing* the distance between said locations by said time delay to determine said relative speed.

^* The invention also provides apparatus for determining a length measurement of a linearly moving material which incorporates a speed determining apparatus as defined above for determining instantaneous speed values of the material over a time interval and means to integrate the instantaneous speed values over said time interval.

The detectors may be radiation detectors to receive electromagnetic radiation reflected and/or emitted from the material at said locations. Such electromagnetic radiation may be in the visible or infra-red spectrum.

The detectors may each comprise a radiation sensor to receive radiation from the respective location, an optical fibre transmitter to transmit the received radiation away from the sensor and photo¬ electric means to receive radiation transmitted by the optical fibre transmitter and to .derive electrical signals therefrom.

Each radiation sensor may comprise a lens system to focus the received radiation on the respective end of the optical fibre transmitter.

The two sensors of the detectors may be incorporated in a single sensing head and the photoelectric means may be installed at a remote location and connected^~to the sensing head by the respective optical fibre trans¬ mitters. BRIEF DESCRIPTION OF DRAWINGS

Application of the invention to the measurement of speed and length of rolled steel product will be explained in some detail with reference to the accompanying drawings in which:- Figure 1 is a perspective view of a speed measuring apparatus constructed in accordance with the invention;

Figure 2 is diagrammatic view of part of the apparatus shown in Figure 1; Figure 3 illustrates the apparatus diagrammatic- ally including certain electrical and signal processing components;

Figure 4 shows the nature of electrical signals derived from two detectors incorporated in the apparatus; Figure 5 is a flow-chart of operations carried out in the signal processing part of the apparatus;

Figure 6 illustrates in block form an alternative signal processing system which may be incorporated in the apparatus of the invention; Figure 7 is assembled-from Figures 7A to 7F* as indicated to form a detailed circuit diagram^*of the signal processing system illustrated in Figure.6; Figure 8 is a flow chart illustrating the operation of the alternative signal processing system; and

Figure 9 illustrates diagrammatically a length measurement system incorporating speed measuring apparatus in accordance with the invention. BEST MODE OF CARRYING OUT INVENTION Figures 1 to 3 illustrate speed measuring apparatus which generally comprises a sensor head 11 connected to a correlation unit 12 by optical fibre cables 13. Correlation unit 12 is connected to a speed display unit 14 and a cathode ray tube display unit 15 can also be provided for initial setting up purposes.

Sensing head 11 comprises an outer cylindrical casing 10. The front end of this casing has two spaced windows 16 and the optical fibre cables 13 extend into the interior of the casing through its back end. The ends of the optical fibre cables within the sensor casing are polished and the casing houses two lens systems each comprising an achromatic doublet lens 17, a collimating aperture plate 18 and a focussing lens 19. When the sensing head is correctly positioned close to a moving strip of material 21, the speed of which is to be measured, the two lens systems receive radiation from two small areas of the product and focus that radiation on to the polished ends of the optical fibre cables. The optical fibre cables may be comprised of fibre bundles of a glass or quartz material so as ef¬ ficiently to transmit the sensed radiation to the correlation unit 12. Within the correlation unit 12, the ends of the optical fibre cables 13 are terminated at a pair of semi-conductor photodetectors 22 producing electrical outputs which are processed by processing circuitry within the correlation unit to produce speed measurements. These speed measurements are displayed digitally on a panel 20 on the front face of unit 12 and also on the remote display unit 14.

In the case where the moving product 21 is a hot rolled steel product, detectable visible and infra-red radiation will be emitted from the surface of the product and the emission will vary from point to point on the product due to variations in surface condition, even on nominally smooth material. In the case of a product which is not hot enough to emit useable amounts of visible or infra-red radiation, an external source of illumination may be trained onto the areas viewed by the sensor head to provide reflected radiation signals which will vary as the material passes through the field of view due to shape or reflectance variations on its surface.

Photodetectors 22 produce electrical outputs each consisting of a high level signal representing the aver¬ age value of the emitted or reflected surface radiation on which there is superimposed a low level fluctuation due to the surface radiation variations on the material passing through the field of view of the sensor. The nature of these signals is illustrated in Figure 4 the upper part of which illustrates the electrical signals derived from the upstream detector (with reference to the direction of movement of the material) and the

OMPI lower part of which shows the electrical output of the downstream sensor. The two outputs are virtually identical but the output of the downstream sensor is displaced in time relative to the upstream output by a time delay indicated as Δt on the Figure. This time delay corresponds to the time taken for travel of the product between the two sensor locations. If the spacing between these two locations (d) is accurately known and the time delay (Δt) can be measured then the speed (v) of the product can be determined by the relationship, d v = —

Δt -

A known method for determining the "time relationship between two signals is by calculation of the cross-correlation of the two signals. The cross-correlation (R ) of two signals is defined 'by the equation below,

T

R τ { x(t)y(t + τ)dt xy T^l→ⁱ-^m∞

where, x(t) and y(t) are the upstream and downstream detector signals in the invention, and τ is an adjustable time delay. If τ is made to vary over a range encompassing the time delay (Δt) between the two sensors, the Rxy will exhibit a maximum value at the delay time τ = Δt, i.e., ^'the signal at the downstream detector (y) .. will most closely resemble the^'^signal at the upstream detector (x) when the signal from (y) is examined a time Δt later than that at which the signal from

(x) is examined.

OMPI Correlation unit 12 incorporates a micro¬ processor for the purpose of cross-correlation of the outputs from the two photodetectors and the signals from the two photodetectors need to be converted to a digital format to be compatible with the microprocessor. The outputs of the photodetectors are therefore individually amplified by means of amplifiers 23 and the amplified signals are then passed through conventional electronic circuitry 24 to produce "one-bit digitization" of the two input signals. Circuitry 24 may comprise low pass filters to remove the high level average radiation signal and two Schmitt triggers the output of each of which is a digital "high" state if the input signal is above a preset threshold level and a digital "low" state if the signal is below the threshold. The resulting digital signals are fed to the microprocessor system 25, which may be based on an INTEL type 8085 microprocessor. As the speed measurement process relies on the accurate determination of a time difference (Δt) it is important to retain an accurately known and fixed time relationship between the sequential samples to be stored in the computer memory for subsequent correlation processing. This is readily achieved by utilizing the Direct Memory Access (DMA) feature of the microprocessor in which the microprocessor ceases all other operations in order to acquire and store samples of both input signals, as and precisely when commanded by an external "clock" signal. This clock signal is derived from a high stability crystal controlled oscillator and therefore consists of a series of pulses which are very precisely spaced in time. Furthermore, the clock pulses from the crystal controlled oscillator can be divided in frequency by external digital circuitry and usually in steps of division by a factor of two in order to ensure that the range of the correlation delay (τ) encompasses the delay time (Δt) . The clock division ratio is selected by the operator of the instrument on the basis of tabulated values of expected product speed versus range selector setting. In practice the selection of the range (or division ratio)^' is not critical as a very wide range of speeds is catered for by each range.

The means of realizing the correlation function will now be described.

The microprocessor recalls from memory the ^• value of the first upstream signal stored under DMA,* and compares this with the first downstream signal, also stored previously under DMA. , This corresponds to a time delay of τ = 0. If the two data bits are identical a digital counter corresponding to τ -= 0 is incremented by one, if they are different the counter is decremented by one, provided that the counter content was previously greater than one. Next, the first ^* upstream data bit is compared with the second downstream data bit resulting in a counter representing a time delay of one clock cyle being incremented or decremented. This process is repeated with successively delayed samples of the downstream data and corresponding . incrementing or decrementing of correlation counters until the maximum delay of interest is achieved. For the purpose of description, it may be assumed that the maximum, delay corresponds to 256 clock cycles, but any number could be chosen.

The above process is then repeated with the next- successive upstream data bit and so on, thus- causing the set of 256 correlation counters to contin- ually increment and decrement. Now, as this is a statistical process and as the input data are

OMPI derived from random variations of surface condition and also as the mean radiation level has been removed by low pass filtering, then it would be expected that the number of data matches should approximately ^* equal the number of data mismatched for all time delays except those around the area corresponding to a time delay of Δt. In other words, most of the correlation counters will exhibit small nett increments or decrements but.those around τ = Δt will exhibit an ever increasing positive count. Eventually one counter will reach a preset threshold value thereby causing a break in the correlation calculation in order that the microprocessor can interrogate the corre¬ lation counters to determine the number of clock cycles of delay which correspond to the maximum or over¬ flowed correlation peak.

The- microprocessor also has been fed with the separation distance (d) of the two detectors, and is able to interrogate the range switch, thus determining the clock rate and hence the actual time delay corresponding to the selected correlation counter. Appropriate division and scaling routines built into the microprocessor then perform the calculation of speed (v) , as described earlier, and display the data on the digital LED display 2.0. A digital to analogue converter (DAC) 26 is provided to produce an analogue output so that the correlation unit can produce both digital and analogue voltage or current representations of the speed signal for remote display and control purposes. The remote display unit 14 may be a digital voltmeter (DVM) connected to the DAC 26.

At this stage the output data and displays are latched and the system is reset to commence a new updated correlation- calculation with one major difference however. The calculation of a 256 clock the prev ously eterm ne pea \ ocat on. T s wenty clock cycle wide "window" tracks^product speed changes by being updated each time a peak is detected and the display updated. This window effect reduces the calculation time, and hence, increases the update rate, by about a factor of ten. A further improvement is achieved by raising the start level or lowering the threshold level of the correlation counters. If the peak is not located in a given number of data sweeps it is deemed to have lost track of the peak. The window is then widened to the full 256 clock cycle width until a peak is relocated.

Programming of the microprocessor to carry out the above described operations is summarized in the software flow-chart of Figure 5.

In the above described apparatus the correlation between the two sets^' of signals derived from the two photodetectors is carried out by the software of the microprocessor. It has been found that this method of operation is quite satisfactory in many practical applications but it is possible to achieve a much higher rate of calculation and update rate by using an integrated circuit or dedicated hardware device for performing the basic correlation process. Such a modification will now be described with, reference to Figures 6, 7 and 8.

Figure 6 is a block diagram of the basic func- tional elements of the hardware correlator and Figure 7 is a detailed circuit diagram which includes additional elements providing.the necessary interface to the controlling microprocessor which, as in the previous case, may be an INTEL-.type 8085 microprocessor.

OMPI ProO The basic elements of the hardware correlator include a ten-bit address counter (ϋ2, U3, U4) ; a 10 + 10 bit summer (U10, Ull, U12) ; two 1 K x four-bit static random access memories (U7, US); an eight-bit correlation counter (U14, U15) ; a buffer (U6) ; and an exclusive OR gate 30. The two memories (U7, U8) are connected so as to allow storage of eight-bit data from the microprocessor, but only one data line in each of the four-bit wide memories is used to collect digitized data from the product.

Addressing of the memories by the microprocessor is^' achieved via the preloading facility of the ten-bit counter (U2, U3, U4) . Data is accessed via the bi- directional buffer (U6) from the memory and a uni¬ directional buffer (U9) from the correlation counter (U14, U15) .

U5 is a latching facility which is used to display information on an oscilloscope either directly or via a digital to analogue converter^' and the other integrated circuits Ul, U6, U9• and U17 shown in the circuit diagram are provided to interface with the microprocessor.

An important feature of the correlator is the split memory architecture. The address on one memory is the output of the ten-bit address counter (U2, U3, U4) while that on the other memory is the sum of this address value plus a ten-bit delay which is selected by the microprocessor. This value appears on the output of the summer (U10, Ull, U12) . As a result of this different address value the data contained in the second memory can be shifted with respect to the first merely by changing the delay input.

During the acquisition of digitized data from the product the buffer shown on the block diagram is enabled thus permitting data to be written into the two memories; upstream data into U7 and downstream data into U8. The rate at which data is written into the memories needsto be precisely fixed'and is controlled by the clock input whose frequency is a function of the lens spacing, expected speed of the product and the number of samples required which in our case is 512. A typical clock rate for 30 mm lens spacing and an expected speed of 2 m/s is 12 KHz. This frequency can be changed in steps to allow for different speed ranges. The collection of data in this manner is a standard direct memory access (DMA) technique. The number of samples collected may be controlled by a programmable counter. After collection the data contained in memory may appear as follows:-

DELAY _ ₀ INPUT

ADDRE^SS 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16...511 COUNTER

UPSTREAM _{1 0 1 0} 0 1 0 0 0 0 1 1 1 . . . .

MEMORY

DOWNSTREAM ₀ 1 1 1 1 0 1 0 0 1 0 MEMORY TRUE j

I DELAY I

The degree of correlation of the upstream and downstream data at a particular time delay is obtained by setting the delay input and enabling the circuit to perform 255 successive reading accesses from the memories. The read clock input_ is not .required to be the same as the sampling frequency and is set to a

\ maximum to minimize processing time. The two-bits of data- (one-bit from each, memory) read on each access are compared by a 2-input exclusive OR logic gate and a match result is used to increment the eight-bit correlation counter. Once this processing sequence is complete the correlation result is obtained by reading the contents of the eight-bit correlation counter. For example, at a delay setting of 4 the above data, if read out of the memories, would appear as follows:-

DELAY _ ₄ INPUT

ADDRESS o 1 2 3 4 5 6 7 8 9 10 11 12 13 14 ^'15 16 ...etc,

COUNTER

UPSTREAM MEMORY 1 0 1 0 0 1 0 0 0 0 1 1 1

^D0WNSTREAM ι 0 1 0 0 1 0 0 0 0 1 1 1 .^'

MEMORY

This delay setting is the peak correlation delay and the two data streams are similar. Hence the correlation counter will exhibit a clear maximum correlation value for this particular time delay.

The hardware is first programmed to obtain 512 two-bit samples of digitized data at an appropriate clock rate and a delay setting of zero. Once the sampled data is obtained the hardware is reprogrammed to perform correlation on 255 successive samples of data. The delay is initially set to zero and the hardware enabled. Once complete the correlation result is obtained, the delay setting incremented and the hardware again enabled. . This process is repeated up to and including a delay setting of 255. periods and the time delay giving the peak correlation value is then determined. This peak correlation delay corresponds to the time for a point on the product to travel between the upstream and downstream sensors. Programming of the microprocessor to carry out the appropriate sequence of operations controlling the functioning of the hardware correlation is shown in the software flow-chart of Figure 8. The main distinction between the software correlation technique described earlier and the operation of the hardware correlator is that in the former case one upstream signal was compared successively with a , series of successively delayed downstream signals and then this process was repeated for successive upstream signals whereas in the hardware correlator a group of successive upstream signals is compared with a corres¬ ponding group of downstream signals and this process is then repeated at increasing delay intervals between the upstream group and the downstream group of signals. In the hardware correlator, the time interval taken by a point on the product to travel between the upstream and downstream sensors is determined by finding the delay corresponding to a maximum correlation count or value and is not dependent on a correlation counter overflow as in the case for the software correlation. Hence, a faster update rate of speed measurement can be achieved.

In the illustrated equipment, the sensing head contains no electrical or electronic components. The optical fibre cables may be quite long so that the signal processing and other electrical components can be removed to a remote location away from any harsh environment, as well as enabling easy .access for periodic checking and maintenance. Calibration of the unit is achieved by two means,

OMPI - -17- - firstly by checking the accuracy of the crystal controlled clock frequency and secondly by checking the alignment of the optical head. The optical head is easily checked, by projecting a bright light source in the reverse direction along the optical fibres such that two small light spots are created on the product or at a position where the product would be expected. The spacing and orientation of these two light spots is easily checked by standard mechanical or electronic means.

Figure 9 illustrates diagrammatically a system for measuring the length of the moving product using the speed measuring apparatus of the present invention. The time taken for the speed measuring unit to stabilise on a^' reliable reading (lock-in time) is compensated for by allowing the material to travel a known distance d, past the sensing head before the speed readings are integrated for the time the product is present in order to calculate the length of the material. The integration process is a known summing technique performed by the microprocessor system and is started and stopped by product presence or absence. The length of the product is calculated by adding the integrated speed values to the known distance d, and outputting this result to a * digital panel meter or in a form suitable for control purposes. These control purposes might be shear control or rerouting of under- or over-length product.

Another known method of non-contact speed and length measurement involves the determination of the Doppler frequency shift experienced by a laser beam reflected from a moving product. The correlation unit described here has the advantage over the Doppler technique that, for hot product, no source of external illumination is required, and for cold product, even though a source of illumination is required, it need

OMPI not be a coherent source such- as a laser beam, a d.c. driven incandescent or fluorescent source will suffice.

INDUSTRIAL APPLICABILITY Apparatus according to the invention may be applied to the measurement of speed and/or length measurements of a wide range of material in solid, liquid or slurry form. It has particular application in the steel industry for measuring the speed and length of shee^*έ,strip and wire products. In other industries, the apparatus would be particularly suitable for use with woven or printed materials, it could also be mounted on a moving vehicle and trained on a track or road ^'. surface to determine the speed of the vehicle and/or a distance travelled.

OMPI O .

Claims (17)

-19- - - - - ■ - - .. ..CLAIMS

1. A method of determining a relative speed between two bodies comprising deriving from a pair of detectors fixed relative to one of said bodies first and second time sequences of signals characteristic of physical variations in the other body at two locations spaced apart in the direction of movement of said material; determining the time delay between the two sequences of signals for which the two sequences most closely match one another; and dividing the distance between said locations by said time delay to determine the speed of the material.

2. A method as claimed in claim 1, wherein the two time sequences of signals derived from the detectors are initially in the form of continuously fluctuating electrical signals and said time delay is determined by the following steps: converting the two sequences of electrical signals to digital form; storing the two sequences of signals in digital form; making separate comparisons of each signal of the first sequence with digital signals of the_ second sequence derived at differing time lags behind said signal of the first sequence; and determining which of said differing time lags produces maximum correlation between the compared signals of the first and second sequences.

3. A method as claimed in claim 2, wherein the making of the separate comparisons of each signal of the first sequence with signals of the second sequence is carried out by comparing a first digital signal of the first sequence successively with digital signals of the second sequence derived at successively greater time lags behind said first digital signal; and comparing successively later signals of the first sequence each successively with digital signals of the second sequence derived at successively greater time lags behind said later signal of the first sequence.

4. A method as claimed in claim 2, wherein the making of the separate comparisons of each signal of the first sequence with digital signals of the second sequence is carried out by comparing a group of successive signals of the. first sequence successively with the groups of success¬ ive signals of the second sequence corresponding to successively greater time lags behind the group of* the first sequence.

5. A method as claimed in claim 4, wherein the determination of the time lags for maximum correlation is carried out by counting the number of matches between the signals of the group of the first sequence and the signals of the comparison group of the second sequence for each of the time lags and determining which time lag produces the maximum number of matches.

6. A method for determining a length measurement of a linearly moving material which comprises determin¬ ing by the method defined in any one of the preceding claims instantaneous material speed values throughout a time interval during which the material, passes the detectors and integrating the instantaneous speed values over said time interval.

7. A method as claimed in claim 6, wherein the material is allowed to move through a particular known distance prior to the time interval for which the instantaneous speed values are integrated and the length measurement is derived by addition of the known distance with the integrated speed values.

8. Apparatus for determining a relative speed between that apparatus and another body, comprising a pair of detectors to derive first and second time sequences of signals characteristic of physical variations in said other body at two locations spaced apart in the direction of said movement and signal processing means for determining the time delay between the two sequences of signals for which the two sequences most closely match one another and for dividing the distance between said locations by said time delay to determine said relative speed.

9. Apparatus as claimed in claim 8, wherein the detectors are such as to derive the two time sequences of signals in the form of continuously fluctuating electrical signals and wherein the signal processing means comprises:- means to convert the two sequences- of electrical signals to digital form; signal storage means to store the two sequences of signals in digital form; and signal processing means operable to make separate comparisons of each signal of the first sequence with digital signals of the second sequence derived at differing time lags behind said signal of the first sequence and to determine which of said differing time lags produces maximum correlation between the compared signals of the first and second sequences.

10. Apparatus as claimed in claim 9, wherein said signal processing means is operable to compare^' a first digital signal of the first sequence successively with digital signals of the second sequence derived at successively greater time lags behind said first digital signal and to compare successively later signals of the first sequence each successively with^" digital signals of the second sequence derived at successively greater time lags behind the later signals of the first sequence.

11. Apparatus as claimed in claim 10, wherein the signal processing means comprises a series of correlation counters to receive the results of the signal comparisons at said differing time lags such that each correlation counter receives the result of all signal comparisons for one particular time lag.

12. Apparatus as claimed in claim 9, wherein the signal processing means is operable to^•compare a group of successive signals of the first sequence successively with the groups of successive signals of the second sequence corresponding to successively greater time lags behind the group of the first sequence.

13. Apparatus as claimed in claim 12, wherein the means to determine the time lag for maximum correlation comprises means to count the number of matches

OMPI between the signals of the group of the first sequence and the signals of the group of the second sequence for each of the successively greater time lags and to determine which-time lag produces the maximum number of ma±ches.

14. Apparatus as claimed in any one of claims 8 to 13, wherein the detectors comprise a radiation sensor to receive radiation from the respective location, an optical fibre transmitter to transmit the received -radiation away from the sensor and photoelectric means to receive radiation transmitted by the optical fibre transmitter and to derive electrical signals therefrom.

15. Apparatus as claimed in claim 14, wherein each radiation sensor comprises a lens system to focus the received radiation on the respective end of the optical fibre transmitter.

16. Apparatus as claimed in claim 14.or claim 15, wherein the two sensors of the detectors are incorporated in a single sensing head to be located in the vicinity of the moving material and the photo¬ electric means and data processing means are mounted in a separate unit to which the optical fibre trans¬ mitters extend from the sensing head.

17. Apparatus for determining a length measurement of a linearly moving material which incorporates a speed determining apparatus as claimed in any one of claims

8 to 16 for determining instantaneous speed values of the material over a time interval and means to integrate the instantaneous speed values over said time interval.

OMPI