The present invention relates to a measuring device, equipment and method for monitoring the onset of corrosion affecting steel reinforcements embedded in concrete, particularly but not exclusively corrosion affecting prestressing cables or steel bars embedded in concrete, according to Claims 1, 17 and 22 respectively. The problem of localized corrosion affecting steel prestressing cables and non-tensioned steel bars forming part of a steel reinforcement embedded in concrete is well known, and rhas fundamental implications for safety, ' especially in road engineering works such as motorway bridges or viaducts. The penetration of chlorides into concrete can lead to the onset of localized corrosion, known as pitting corrosion, which is followed by acidification of the closed cell with the formation of hydrogen in the cavity in which corrosion is taking place. The hydrogen also causes the embrittlement of the steel used for these reinforcements. Because of this process, therefore, whenever corrosion is initiated, this initiation can lead to a catastrophic fracture of the steel cables without any obvious or appreciable warning, with repercussions on safety which can easily be imagined. There are known methods of measuring the electrochemical potential, in which the onset or propagation of corrosion in reinforced concrete structures exposed to atmospheric agents is monitored by potential mapping. In particular, this mapping is based on the electrochemical mechanism of localized corrosion, in which the anodic corrosion areas, which are small, are separated from the surrounding cathodic areas, with an exchange of electrical current between them. More precisely, the anodic areas are at a more negative potential than the cathodic areas, and the anodic corrosion areas can therefore be identified accurately by measuring the potential at each point, using a grid of suitable size. For this purpose, there are international standards which can be used to determine whether or not a corrosive phenomenon is occurring in a reinforcing bar in concrete exposed to the atmosphere. Indeed, the interpretation of the data is covered by both international and national standards and regulations, such as ASTM C 876 and UNI 9535. For example, the ASTM C 876 standard specifies that the areas in which corrosion is occurring show a potential which is more negative than the potential measured by a reference electrode, while areas not subject to corrosion, or passive areas, show a potential which is more positive than that measured by the same reference electrode. For example, if a copper-copper sulphate reference electrode, commonly known as CSE, is used, it will be found that the areas in which corrosion is occurring show a potential of less than -0.35 V CSE, while areas not subject to corrosion show a potential of more than approximately -0.20 V CSE. In other words, if the potential measured by the electrode falls below -0.35 V CSE there is a high probability of corrosion in one or more points of the reinforcement. However, the aforesaid potential mapping method is not applicable to prestressing cables, since these cables are positioned inside a metal or polymer sheath, and are therefore insulated from the surrounding environment. Furthermore, the technology available at present cannot be used to detect the onset of corrosion in either the initiation or the propagation stage. It is therefore necessary to make use of indirect signal detection, possibly before the start of propagation of the fracture. It should be noted that, for the purposes of safety, it is not necessary to know the exact point at which corrosion has been initiated, but it is essential to know whether or not the corrosion phenomenon has occurred. This is because the cable, or the structure, must be replaced as soon as corrosion occurs . US Patent 5,403,550 proposes the use of a continuous wire electrode to determine localized corrosion in steel reinforcing cables embedded in concrete. In particular, it proposes the use of a single wire electrode extending along the longitudinal development of the reinforcing cables (or bars) and in the proximity of the said cables, in order to detect the initiation of corrosion at any point. However, an electrode of this type can supply a signal indicating the onset of a localized attack for certain geometries only, namely for geometries with an "1/D" ratio of less than approximately twenty, where "1" denotes the linear dimension of the wire electrode and of the reinforcing cable, and "D" denotes the thickness of concrete (mortar or cement grout) between the strand and the outer insulating sheath, or, in the case of a bar, between the bar and the outer surface of the concrete exposed to atmospheric agents. It has also been found experimentally that the potential measured by means of the continuous wire electrode, as described in US patent 5,403,550, is equal to the weighted mean of the potentials of the corresponding surfaces. This finding can be explained by the fact that the surface of the strands at the point of corrosion (pitting) is much smaller than the rest of the surface, which remains in the passive state. In these conditions, the surface which is the point of corrosion is smaller, in percentage terms, than 0.1% of the total surface area of the strand. The above findings all confirm that the weighted mean of the potentials is practically coincident with the value of the passive surfaces which are not subject to corrosion, since, for real structures, the ratio "1/D" is greater than a thousand on average. This means that the localized corrosion at a given point of the cable is not in fact detectable by measuring the potential with an electrode of this type. In view of the state of the art described above, the object of the present invention is to provide a device which can overcome the limitations of the known art which has been described. A further object is to provide equipment for monitoring the onset of localized corrosion affecting the steel cables embedded in reinforced concrete. Finally, another object of the present invention is to provide a method for monitoring the onset of localized corrosion affecting the steel cables and bars embedded in reinforced concrete. According to the present invention, this object is achieved by means of a device for measuring the electrochemical potential of a steel cable or bar in concrete as claimed in Claim, 1, and this object is also achieved by means of equipment for measuring the electrochemical potential of a steel cable or bar in concrete as claimed in Claim 17, and finally it is achieved by means of a method for measuring the electrochemical potential of a steel cable or bar in concrete as claimed in Claim 22. The present invention makes it possible to carry out continuous monitoring of the localized corrosion conditions affecting steel cables embedded in prestressed reinforced concrete or affecting non- tensioned bars, in other words the reinforcements normally used in bridge slabs, bearing blocks, pillars, reinforced poles, and the like. The characteristics and advantages of the present invention will be made clear by the following detailed description of a practical embodiment, illustrated by way of example and without restrictive intent in the attached drawings, in which: - Figure 1 shows a wire electrode for measuring the electrochemical potential according to the present invention; Figure 2 shows a measuring device according to the present invention; - Figure 2A shows a sectional view taken along the line II-II of Figure 2; Figure 3 shows a sectional view of a generic prestressing cable according to the prior art; Figure 4 shows a schematic view of the measuring electrode of Figure 2 after it has been applied to the strands of a prestressing cable; Figure 5 shows a schematic view of equipment for monitoring the localized corrosion according to the present invention; - Figure 5A shows a sectional view taken along the line III-III of Figure 5; Figure 6 shows a schematic view of the measuring electrode of Figure 2 after it has been positioned on the bars of a reinforced concrete structure. With reference to Figure 1, this shows a wire electrode 1 comprising an active portion 2, associated for operation with a transmission means 3. The active portion 2 of the electrode 1 is the transducer element which can detect a physical magnitude such as potential, while the transmission portion 3 acts as a means of carrying the information obtained by the active portion. The active portion 2 takes the form of a wire consisting of a metallic material, preferably selected from the group comprising iron, stainless steel, hot-oxidized stainless steel, titanium activated with noble metal oxides, nickel, and copper. Advantageously, the metallic materials are copper and stainless steel, since these are easily obtainable on the market. The wire or active portion 2 of the wire electrode 1 has a linear length "L", this length "L" being a function of the specific applications, as described below. The transmission means 3 may, for example, take the form of an electrical cable. For example, if copper is selected as the metallic material to form the active portion 2, the cable 3 can be ordinary commercially available electrical cable, used for constructing electrical installations for public and/or industrial use. The junction between the active portion 2 and the transmission means 3 is made by processes known to those skilled in the art; alternatively, according to an optional aspect of the present invention, the active portion 2 and the transmission means 3 are connected to each other by connecting means 4. These connecting means 4 connect the active part 2 electrically to the transmission part 3, while also acting means of insulation from the concrete. This prevents the concrete from penetrating into the junction area between the wire 2 and the electrical cable 3. These connecting means 4 may, for example, take the form of a heat-shrinking sheath 4A, acting as the electrical connection element, wrapped with a sealing resin 4B which acts as the insulating element. A measuring device 5 according to the present invention is shown in Figure 2, which shows that the said measuring device 5 comprises a plurality of wire electrodes 1, indicated respectively by IA, IB, 1C, etc., each of which consists of its own electrical connecting cable 3 and active portion 2. In turn, this measuring device 5 is contained (or wrapped) in an insulating sheath 7 which can resist the alkalinity of the concrete. The sheath 7 is made, for example, from a perforated plastics material. The measuring device 5 thus resembles a single multistrand cable. Figure 3 shows a section through a prestressing cable 8 forming part of a steel reinforcement. The cable 8 is formed, for example, from a core consisting of a plurality of strands 9, the latter being contained (or wrapped) in a sheath 10. The strands 9 can be, for example, interleaved and wound together in spiral form. Between the sheath 10 and the strands 9 there is a gap (or interstice) "D", called the concrete cover. This space "D" is filled with alkaline mortar or cement grout (which acts as an electrolyte) . The measuring device 5 has to be inserted into the gap formed by the space "D". Moving on to Figure 4, it will be noted that the measuring device 5 is fitted on top of the strands 9 of a prestressing cable 8. This measuring device 5 can be positioned in the proximity of the said strands 9 and/or directly associated with the said strands 9. In particular, it should be noted that Figure 4 shows a possible arrangement of the wire electrodes IA, IB, 1C, ID along the strands 9 in the case in which there is a direct association between the measuring device 5 and the strands 9 of the cable 8. The wire electrodes IA, IB, 1C and ID, forming the measuring device 5, are positioned in the concrete cover D, in other words in the gap D formed between the strands 9 and the sheath 10. It should also be noted that the wire electrodes IA, IB, 1C and ID are positioned so as to cover a portion of the linear length "1" of the strands 9. It is also possible for the wire electrodes IA, IB, 1C and ID to be positioned so as to cover the whole of the linear length "1" of the strands 9. More precisely, it should be noted that the active portions 2 of each wire electrode are positioned in series with each other. In other words, the terminal part (the tail) of the active portion 2 of the first wire electrode IA terminates at the start of the leading part (the head) of the active portion of the second wire electrode IB, and so on. This results in a "head-to-tail" arrangement. In the specific embodiment shown in Figure 4, the first wire electrode IA has the function of monitoring a first portion of the cable 8, the second wire electrode IB has the function of measuring a second portion of the cable, and so on, the first and second portions each having an equal length "L". In another embodiment, not shown in the figures, it is also possible for the first wire electrode IA to have the function of monitoring one portion of the cable 8, while the second wire electrode IB has the function of measuring another portion of the cable 8, without these portions being consecutive with respect to each other. With reference to Figure 5, it may be noted the measuring device 5 is wound in a spiral with a suitable pitch λλP" around the strands 9 of the cable 8. The pitch "P" of the winding of the said device 5 around the strands 9 is a function of the specific application and is, for example, in the range from one to three metres, or is preferably equal to two metres. In another embodiment, not shown in the figures, the measuring device 5 can be extended parallel to the strands. With particular reference to Figure 6, what is shown here is another possible application of the measuring device 5. In Figure 6, it can be seen that the measuring device 5 is installed in the proximity of the bars 11 of a non-tensioned structure 12 and an outer surface 13. For example, the non-tensioned structure 12 could be of the type used in the slabs of bridges, bearing blocks, pillars, reinforced poles, and the like. In particular, the bars 11 of the non-tensioned structure 12 are embedded in the concrete at a distance λλd" from the outer surface 13. In this embodiment also, the measuring device 5 is to be positioned in this thickness "d". It should be noted that the thickness λλd" is also usually called the "concrete cover". In the specific application described in Figure 6, the measuring device 5 is again installed according to the previous description given with reference to Figure 4. The ratio between the length λλL" of the active portion 2 of each wire electrode IA, IB, 1C, etc. and the concrete cover "D" (or "d") is of fundamental importance for the identification of the onset of localized corrosion along the strands 9 of the cable 8 or along the bars 11 of the non-tensioned structure 12. Thus the ratio "L/D" (or "L/d") can be used to determine the most suitable geometry for monitoring the strands 9 or bars 11 of any steel reinforcement embedded in the concrete. In particular, the interval "L/D", according to the present invention, is in the range from zero to thirty, in other words: L/D = 0 í 30 or preferably 15 í 20. In other words, given the length "1" of the strands 9 of the cable 8 and the available concrete cover "D", this ratio L/D can be used to find the requisite linear length "L" of the active portion 2 of each wire electrode IA, IB, 1C, etc. making up the measuring device 5, and the number of steel wires making up the said device. For example, if the onset of corrosion is to be monitored on strands 9 with a length "1" equal to eighteen metres, and the concrete cover "D" is equal to one centimetre, then, in the case in which the ratio "L/D" is assumed to be thirty, we find that the length λλL" of each active portion 2 of the wire electrode must be thirty centimetres . Thus, in order to cover the whole length "1" of the cable 8, it is necessary to provide sixty individual wire electrodes IA, IB, etc., each of which must be provided with both an active portion 2 and a corresponding electrical signal transmission cable 3. A similar reasoning can be applied to the case of the bars 11 of the concrete structure 12. This is because, given the length "1" of the bars 11 and the available concrete cover "d", the ratio "L/d" can be used to find the requisite linear length λλL" of the active portion 2 of each wire electrode IA, IB, 1C, etc. making up the measuring device 5, and the number of steel wires making up the said device. As mentioned above, the first stage of the corrosion phenomenon is the formation of a first corrosion cell (a pit) , at any point of a steel reinforcement, regardless of whether this consists of a cable 8 or of the bars 11 of the structure 12. The appearance of the pit can be detected in real time by the "local" measurement of the electrochemical corrosion potential of the strand 9 or the bar 11, in other words by a measurement made in the immediate vicinity of the corroded area itself, exploiting the presence of the active portion 2 of each wire electrode IA, IB, 1C, etc. Referring to Figures 5 and 5A, which show equipment for measuring the electrochemical potential in the cable 8, it can be seen that the measuring device 5 is wound in a spiral with a pitch "P" on the strands 9 along the whole linear length "1" of the cable 8. In particular, each wire electrode IA, IB, 1C, etc. forming the measuring device 5 is connected to a junction box or data acquisition card 14. In turn, this data acquisition card 14 is connected to a potential measuring instrument 15, for example a high-impedance voltmeter. Specifically, the transmission cables 3 of each wire 2 are connected to the data acquisition card 14. This card 14 also has the function of housing one or more devices (not shown in the figures) for carrying out processing operations on the measurements made by each wire electrode and/or for storing these measurements and/or the data extrapolated from them. These devices can, for example, take the form of non¬ volatile memories and processors. The voltmeter 15 has the cable 8 as its positive terminal, while its negative terminal is the acquisition card 14, in other words the transmission cables 3 of each active portion 2. If λλi" denotes the number of wire electrodes IA, IB, 1C, etc. used to monitor the strands 9 or a portion of the strands, then when one or more of the said wire electrodes IA, IB, 1C, etc. is in the vicinity of an area in which the corrosion phenomenon is occurring, each of the said wire electrodes IA, IB, 1C, etc. detects its own potential difference δVj.. If even a single measurement of the potential difference δVi is more negative than the preceding measurement of the same wire electrode, and/or more negative than the mean of the other wire electrodes, by a predetermined threshold value ξ, then corrosion has been initiated at this point of the strand. It should be noted that the mean can be calculated on the basis of a measurement made in real time by all the wire electrodes IA, IB, 1C, etc., or by taking a historic mean of the measurements made previously by these wire electrodes IA, IB, 1C, etc., or by a combination of these methods. If one or more of the wire electrodes IA, IB, 1C, etc. supplies a potential measurement below the predetermined threshold value ξ, for example less than at least one hundred millivolts, and in any case below the preceding mean of the said threshold value ξ, in other words at least one hundred millivolts, this means that the phenomenon of localized corrosion is occurring in the area to which the wire electrode relates. According to the present invention, the method by which the potential difference δVi is measured is as follows: a potential difference δVi is measured for each λλi", where "i" is the number of wire electrodes used to monitor the strands 9 of the cable 8 or the bars 11 of the non-tensioned structure 12; a statistical analysis is performed on the measured potentials AV1; the measurement of the potential difference δVi is repeated with a frequency T specified in advance; in other words, the potentials δVi are measured at the time T=I for each "i", the measurement of the potentials δVi are repeated at the time T=2 for each "i", the potentials δVi are repeated at the time T=3 for each "i", and so on; if necessary, a step of storing the said measured data, namely the value of the potential difference δVi at the time T=I for each λλi", the value of the potential difference δVi at the time T=2 for each "i", the value of the potential difference AV1 at the time T=3 for each "i", and so on, is carried out; each measurement is compared with the preceding result and with the mean μ of the preceding measurements, in other words δVi at the time T=3 is compared with the measurement of δVi at the time T=2 and with the mean value μ of the measurements of δVi at the time T=I and at the time T=2; each measurement is compared with the value of the variance σ of each preceding measure, and finally a signal is generated when the measurement of the said potential δVi of only one of the said wire electrodes IA, IB, 1C, etc. yields a value which is lower than the preceding value by at least a predetermined value ξ. The frequency T of repetition of the measurement is a function of the specific application, and can vary from minutes to years. Preferably, the frequency of measurement T is set at one week. In particular, the statistical analysis of the measured values is carried out by determining the minimum and/or maximum and/or mean value of each measurement made and the value of the variance σ of each measurement. Thus the onset of localized corrosion in the strands 9 of prestressing cables 8, or in the bars 11 of reinforcing structures 12, embedded in concrete can be detected in good time by continual measurement of the potential δVi between the measuring device 5 and the cable 8 or the bar 11 of the non-tensioned structure 12. Clearly, a person skilled in the art can make numerous modifications and variations to the configurations described above, in order to meet contingent and specific requirements, all such modifications and variations being contained within the scope of protection of the invention as defined in the following claims.