1、1School of Information and Electrical Engineering, China University of Mining & Technology, Xuzhou, Jiangsu 221008, China2School of Mechanical Electronic and Information Engineering, China University of Mining & Technology, Beijing 100083, ChinaAbstract: In order to overcome the flaws of present dom
2、estic devices for detecting faulty wires such as low precision,low sensitivity and instability, a new instrument for detecting and processing the signal of flux leakage caused by broken wires of coal mine-hoist cables is investigated. The principle of strong magnetic detection was adopted in the equ
3、ipment. Wires were magnetized by a pre-magnetic head to reach magnetization saturation. Our special feature is that the number of flux-gates installed along the circle direction on the wall of sensors is twice as large as the number of strands in the wire cable. Neighboring components are connected
4、in series and the interference on the surface of the wire cable, produced by leakage from the flux field of the wire strands, is efficiently filtered. The sampled signal sequence produced by broken wires, which is characterized by a three-dimensional distribution of the flux-leakage field on the sur
5、face of the wire cable, can be dimensionally condensed and characteristically extracted. A model of a BP neural network is built and the algorithm of the BP neural network is then used to identify the number of broken wires quantitatively. In our research, we used a 637+FC, 24 mm wire cable as our t
6、est object. Randomly several wires were artificially broken and damaged to different degrees. The experiments were carried out 100 times to obtain data for 100 groups from our samples. The data were then entered into the BP neural network and trained. The network was then used to identify a total 16
7、 wires, broken at five different locations. The test data proves that our new device can enhance the precision in detecting broken and damaged wires.Key words: wire cable; broken wire; signal processing; detection deviceCLC number: TB 421 IntroductionIt is well-known that coal mine-hoist cables are
8、an important part in coal mine-hoists or transportation systems. Wires are, in fact, subjected to breakage due to wear, corrosion and fatigue. The extent of damage and the carrying capacity of wires are directly related to the safety of equipment and staff. At present, there are many detection devic
9、es for broken steel cables manufactured in China, but most devices do not meet the conditions ideally required in practice. The reasons are largely the complex structure of wires, bad working conditions, the multiplicity and uncertainty of broken wires. It is therefore quite difficult to detect sign
10、s of broken wires as well as to analyze and process detected signal of broken wires in cables 1.A new instrument for broken wires detection and procession of coal mine-hoist cables was investigatedfor this paper. With the special structure of a detection transducer, the interfering signal from the l
11、eakage field of wire twists can be filtered efficiently. After the extraction of dimensional contraction and characteristic values of multi-ways signals, a quantitative BP neural network recognition for broken wires in steel cables was realized. The test results are presented.2 Basic Structural Prin
12、ciple of the On-Line Detection Instrument for Coal Mine-Hoist CableThe structural principle of the on-line detection device for wire cables studied by us is shown in Fig. 1.The detection transducer is composed of two semicircle cylindrical structures which can be opened or closed. The magnetic sensi
13、ng unit is a fluxgate unit made of a single magnetic core and is single-winding. Some magnetic sensing units are evenly arranged around the inner wall of the transducer, the number of which is twice as many as the number of the wire strands in the inspected cable. As well, two neighboring units are
14、connected in series to a detection channel.Consequently, the number of detection channels of the detection instrument is equal to the number of wire strands in the cable.Fig. 1 Structural principle of detection instrument for broken wires in coal mine-hoist cables.After being filtered and reshaped,
15、the detection signal from each channel is sent to the signal processing unit. The analog detection signal is converted into adiscrete dimensional sequence of sampling values by multi-channel A/D conversion, followed by a characteristic extraction, a BP neural network recognition and the output of th
16、e result. When viewed separately, the leakage field signal detected by each single fluxgate unit is the leakage field intensity in the steel cable where the corresponding fluxgate units are located. That is, the outputsignal Zjk of any jth test unit is:where FC is the structural parameter of the flu
17、xgate, the width of the drive square-wave, s the saturated magneto-conductivity rate, B c, j the magneticinduction intensity of the leakage field produced by broken wires, Br, j the magnetic induction intensity of the leakage field produced by wire cable twists, Zf j the signal value of b
18、roken wires and Z r, j the value of the interference signal produced by wire cable twists.After , F C ，a , us , Fare assured, F is a constant.After the wire cables are deeply magnetized, the numerical value of sis very small. As a result, the value of c, j is larger and there is
19、 no need to magnify and process the detection signal again. When the sensor is operating along wire cables at a specified speed, the signals detected by each of the magnetic fluxgate units can effectively show the three-dimensional distribution status of magnetic flux leakage, generated at the surfa
20、ce of wire cables24.3 Filtration of the Wavelike Oscillation Interference Signal Produced by Cable Wire TwistsThe signal of broken wires from wire cables obtained by a single fluxgate detection unit of the transducer （formula （1） contains all kinds of interfering signals. The effect of the wavelike
21、oscillation magnetic flux leakage B r, j due to the special structure of the steel cables is largest, which directly affects the detection of broken or damaged wires, especially in coal mine-hoist cables. We should consider the possibility of filtering the interference signals. In formula （1）, the i
22、nterference signal r, j caused by a wavelike oscillation shows up as periodic variation. This kind of wavelike oscillation interferencesignal can be regarded approximately as a sine wave,as shown in Fig. 2.Fig. 2 Wavelike oscillation interference signalproduced by the cable twistOver the length dire
23、ction of wire cables, its variation period T is a Lay length of cable wire strands. At the circle direction of the wire cable, its variation period is the reciprocal of the number of outer wire strands of the circle length of the wire cable. Therefore, the wavelike oscillation interference signal of
24、 the jth detection channel can be expresse d as: jwhere a is the Direct Current Component of the wavelike oscillation signal, m the Alternating Current Component magnitude of the wavelike oscillartion signal, T represents the value of periods, y is the position of the detection unit, starting from t
25、he initial spot, j the initial phase of the wavelike oscillation signal, N the number of wire strands of the steel cable, and is the number of detection units. cObviously when c , i.e., when the number of detection units doubles the number of outer strands of the wire cable, the wavelike o
26、scillation signal contained in the leakage magnetic field signal inspected by any two neighboring detection units is in a reversal phase. Therefore, when the neighboring detection units along the inner wall of the cylinder of the transducer structure are connected forward into a test channel in seri
27、es two by two, it is equivalent to adding the （j+1）th test channel signal to the jth test channel signal. Thus the strand peak value of the wavelike oscillation signal compensates for the strand value for the moment. That is, at this moment, the only remaining wavelike oscillation signal is the Dire
28、ct Current ComponenAt this moment, the magnetic field signal of leakage from any of the inspection channels made up of the fluxgate array should be:of this formula can be eliminated when the zero detection position is adjusted. Therefore, we considered that the wavelike oscillation interference sign
29、al of cable wires is filtered by formula （4）. After this pretreatment, each leakage from broken wires, shown by magnetic field signals from the transducer, becomes a channel sample value by A/D conversion, as shown in Fig. 3.Fig. 3 Multi-channel sampling value of broken wiresignals from wire cables4
30、 Extraction of Characteristic Value of Signals from Broken WiresAs is shown in Fig. 3, the N-channel inspection signals from the transducer becomes its sampling sequence by A/D conversion. If the number of samples of the signals of broken wires is K, the sequence of broken wire sample signals of the
31、 jth channel can be expressed as a row vector with K elements.The N-channel signal sequence will make up a N-dimensional series vector group of broken wiresignals:At this moment, Z is a characteristic matrix of broken wires and it contains all the information on the status of the broken wires. N
32、48693;K Given the analysis of repeated experiments, the width of the diffused leakage from the magnetic fieldon the surface of wire cables created by broken wires is not larger than 20 mm. When the speed of the inspected wire cable is 3 m/s and the sampling interval is 1.2 mm, the number of samples K is 16 at most. When the number of inspection channels is N=4, Z should be a 416 matrix. If the analysis of the characteristic matrix of broken or damaged wires Z were directly carried out, the analytical proc
