Reivindicaciones
1. A method for real time tests and diagnosis of the sources of partial discharges produced in high voltage equipment and installations, which are in service or out of service, by means of discriminating background noise, identifying the position of the PDs, distinguishing between different sources of PDs located in a single position, determining, for each source of PD, its magnitude, rate of repetition and type of associated fault, characterized in that it comprises: - synchronised capturing and digitising (block 1.1) of the analogue signals collected by two partial discharge sensors located at the ends of the installation to be diagnosed together with the analogue signals collected from two voltage sensors, one for sensing the test voltage and the other one for sensing the grid voltage, all the captures starting in the instant that the UTC (Universal Time Coordinated) coincides with exact seconds, being supported for such purpose on, for example, the PPS signal (pulse per second) delivered by a GPS receiver, and each capture covering one or more complete periods of the test voltage wave (at least 20 milliseconds for 50 Hz); - segmenting (block 1.2) the sensed and digitalised signal in time intervals equal to the periods of the sine wave test voltage signal. - real time discriminating (block 1.3), through the use of multiprocessors, within the signal measured by each PD sensor captured (block 1.1), between electric noise and the partial discharge-type (PD) signal by means of applying the Wavelet transform together with a statistical treatment of the measured signal; - determining (block 1.4) for each PD-type signal, the time instant, t<i> , in which it occurs on the absolute UTC time reference, associated with the phase difference of the PD-type signal in relation to the zero-crossing of the test voltage wave, φ<i> , and in relation to the zero-crossing of the grid voltage wave, φ<ri> , in the event that the latter is different from the test voltage; as well as the real time determination, through the use of multiprocessors, for each i-th PD-type signal, of the parameters of the mathematical function derived from the hyperbolic secant containing two parameters α<i> and β<i> making it asymmetrical e i i = A i e α i ⋅ t - t 0 i + e - β i ⋅ t - t 0 i achieving the best fit with the envelope of the damped oscillating event characterising the PD-type signal and extracting from this function the two parameters representative of the shape of the envelope, in relation to the rise time and fall time, which are associated with the two coefficients α<i> and β<i> of the denominator of the asymmetrical hyperbolic secant function; together with the real time determination, through the use of multiprocessors, for each PD-type signal, of the parameters of the sinusoidal mathematical function g i t = sine ω i ⋅ t - ψ i which best fits with the damped oscillating event and characterises the PD-type signal; extracting from this function the parameter of the fundamental frequency f<í> =ω<í> /2π as an additional parameter representative of its shape; - storing (block 1.5) the calculated parameters characteristic of each partial discharge pulse determined in block (1.3). - repeating (block 1.6) the stages associated with blocks (1.2) to (1.5) for each period interval of the test voltage (20 ms for 50 Hz and 16.66 ms for 60 Hz) of the global period of the captured PD signal (block 1.1). - repeating new captures (block 1.7) for the purpose of again applying the preceding stages associated with blocks (1.2) to (1.6) enough times to enable having statistically sufficient data for reliably determining the position of sources of the partial discharges, the number of different sources of partial discharge located in each position and the parameters representative of each source of PD, which correspond to the magnitude of the discharge, the rate of repetition of PD pulses and the phase resolved patterns of the test voltage and of the grid voltage; the number of acquisitions considered sufficient for the mentioned calculation is 250 periods of the grid voltage wave; - excluding (block 1.8) PDs originating in the grid voltage far from the element under testing when the grid voltage is different from the test voltage by means of representing the magnitude of the phase resolved discharge of the grid voltage (j)<r> i); - correlating (block 1.9) the parameters of each PD signal measured with the two measurement systems to enable determining the position of sources of PDs by using the absolute time reference obtained in the stage associated with block (1.4); - determining (block 1.10) the position map of sources of PDs taking into account the delay in arrival to each sensor of the PD pulses measured with a different measurement system and paired by proximity, x<i> (Δt<i> ), taking into account the maximum time delay possible between the PD pulses received by the two measurement systems, the ratio between the maximum length to be travelled by the PD pulse and the propagation speed of the PD pulse; - identifying (block 1.11) the different number of sources of PD for each position of sources of PDs through the formation of groups of PD signals having similar values of the three following parameters: the two parameters representative of the shape of the envelope of the PD-type signal (block 1.4), related to the rise time and the fall time α<i> and β<i> and the parameter of the fundamental frequency f<i> of the damped oscillation pulse characteristic of the PD signal; - determining (block 1.12) for each source of PD the value representative of the magnitude of the discharge, of the rate of repetition of PD pulses associated with each acquisition second, as well as the phase resolved pattern of the PD pulses of the test voltage wave, correlating it in relation to typical fault reference patterns, for example by means of a neural network structure trained in known typical fault patterns; - storing (block 1.13) the parameters determined in blocks (1.10), (1.11) and (1.2); - repeating the process (block 1.14) for determining parameters calculated in block (1.12) for each source of PD - repeating the process (1.15) indicated in blocks (1.11) to (1.14) for the sources of PDs associated with each position of PDs. - graphically representing (block 1.16) the results of the stages associated with blocks (1.10), (1.11) and (1.12), corresponding to the position map of PDs where the sources of PDs are located (block 1.10), groups of different sources of PDs located in each position (block 1.11), results of the parameters obtained for each source of PD (block 1.12) corresponding to the magnitude of the discharge, the rate of repetition, the phase resolved patterns of PDs of the test voltage and the result of the type of fault associated with each source of PD of the network. 2. The method according to the method of claim 1, corresponding to the stage of the noise discrimination block (1.3) which allows real time distinction of signals with a transient waveform characteristic from a PD of other signals characteristic of electric background noise superimposed by using multiprocessors working in parallel. 3. The method according to the method of claim 1, corresponding to the stage associated with block (1.4) relating to the real time determination, through the use of multiprocessors, of the parameters of the mathematical function referred to as asymmetrical hyperbolic secant e i i = A i e α i ⋅ t - t 0 i + e - β i ⋅ t - t 0 i which best fits with the envelope of the damped oscillating event characterising each PD-type signal sensed in the stage associated with block (1.1) and of the parameters of the sinusoidal function g i t = sine ω i ⋅ t - ψ i which best fits with the frequency and phase difference of the damped oscillating event characterising each PD-type signal sensed in the stage associated with block (1.1), extracting from the first function the two parameters representative of the shape of the envelope, in relation to the rise time and fall time, which correspond to the two coefficients α<i> and β<i> of the exponentials of the denominator of the asymmetrical hyperbolic secant function and extracting from the second function the parameter of the fundamental frequency f<i> =ω<i> /2<π> as an additional parameter representative of its shape. 4. The method according to the method of claim 1, corresponding to the stage associated with block (1.8) relating to the exclusion of PDs originated by the grid voltage, far from the element or installation under testing when the test voltage is different from the grid voltage, through the representation of the magnitudes of the phase resolved partial discharges of the grid voltage (φ<i> ) determined in the stage associated with block (1.4). 5. The method according to the method of claim 1, corresponding to the stage associated with block (1.11) relating to the identification of the different number of sources of PDs, through the formation of groups of PD signals having similar values of the three following parameters: the two parameters representative of the shape of the envelope of the PD-type signal related to the rise time and the fall time α<i> , β<i> and the parameter of the fundamental frequency, f<i> , of the damped oscillation pulse characteristic of the PD signal, all of which are mentioned in claim 3. 6. The method according to the method of claim 1, corresponding to the stage associated with block (1.16) relating to the graphic representation of the results of the analysis, stages corresponding to the blocks (1.10), (1.11), (1.12) relating to the position map of PDs where the sources of PDs are located (block 1.10), the grouping of the different sources of PDs located in each position (block 1.11) and the results obtained from each source of PD (1.12): magnitude of the discharge, rate of repetition, phase resolved pattern of the voltage for each source of PD and result of the type of fault associated with each source of PD. 7. A physical system for detecting partial discharge-type (PD) signals based on the method of claim 1 which allows taking simultaneous and synchronised measurements in two measurement systems for measuring PDs, discriminating the noise in relation to the transient waveform characteristic of the PD signals according to the method of claim 2, determining the parameters associated with each PD according to the method of claim 3, determining the map of sources of PDs along the length of the cable according to the stage associated with block (1.10) in claim 1, graphically representing the groups of the sources of PDs according to the method of claim 5, graphically representing the results of the analysis according to the method of claim 6, made up of two independent measurement systems each having the following elements: - a partial discharge sensor; - a voltage sensor for measuring the test voltage waveform; - a voltage sensor for measuring the grid voltage waveform in the event that it is different from the test voltage; - a card for receiving the UTC time signal, for example from a GPS system, and another one for synchronised trigger pulse generation (TPG); - a digital recorder triggered by the synchronisation pulse coming from the synchronised trigger pulse card; - a protection and control equipment, protecting the two digital recorders of the measurement system against surges; - a personal computer with multiple processing capacity through one or several multiprocessing units; - a portable computer for remotely controlling the measurement system. 8. A physical system for generating alternating high voltages of the frequency of the supply voltage (50 Hz or 60 Hz) generating the test voltage during short time periods, in which the PD signals are captured according to the stage associated with block (1.1) of claim 1; after which the voltage is reduced, for a considerably longer wait time period, to a low enough value to allow the dissipation of the heat produced in the instants of generating alternating high voltage for testing, using each reduced voltage period for analysing the measurements relating to the stages associated with blocks (1.2) to (1.7) of claim 1 and thus being able to supply a pulsed state power that is much greater than the steady-state nominal power of the generator, while at the same time the reactive power required in the test is compensated by means of FACT-type power electronics for the purpose of limiting the use of reactors and condensers for compensating the reactive power required in the test.
