LASER WELDING STATION AND PROCEDURE FOR WELDING TOGETHER ELEMENTS OF METAL PRODUCTS

Description

“LASER WELDING STATION AND WELDING PROCEDURE SET OF ELEMENTS OF METAL ARTIFACTS” SCOPE OF APPLICATION The present invention refers to a laser welding station for welding 5 together at least two elements of a metal artefact and a related laser welding process. In particular, the welding station and the related procedure are configured to monitor the reflected light of the artifact due to the incidence of the laser beam responsible for the melting of the metallic material, in manufacturing applications such as welding of elements 10 of electrochemical battery, welding of electrical terminals of a winding or similar. STATE OF THE ART In the field of welding for applications on products made of materials metallic, the use of laser sources configured to transmit is known 15 quickly a welding energy to the elements to be welded together, in order to obtain a good compromise between the quality of the product, plant productivity and welding process cost. In some cases, it is required that the welding be done in such a way as to join stably two elements with a high resistance to tearing force. 20 This mechanical property requirement is achieved thanks to the ability to penetration of the laser beam into the manufactured products during the manufacturing process welding. In fact, there is a specific way of laser welding, which is typically known as keyhole welding. In these cases, the high power density of the laser source allows 25 generate a beam with a very small incidence area on the artefact to be work, in contrast to the heat conduction welding way, which is characterized by having a higher power and energy density low, resulting in a substantially shallow and broader fusion compared to the depth welding method. The thermal welding method, on the other hand, 5 is pursued in cases where a rather superficial welding is required, for example when two thin layers of material are superimposed with the first layer placed between the laser source and the second layer to be welded reciprocally to the first. Returning to the depth or keyhole welding mode, instead, the laser beam 10 accident generates a narrow profile, a real excavation in the material for via the vaporization of metallic material. This fact allows the beam laser to penetrate deep into the material to be welded. At that point, the excavation (keyhole) is surrounded by molten material generated in the meantime due to the Heat transfer between the laser beam and the metal material. This molten material 15 tends to fill the space dug as the laser beam is moved along a welding path in the workpiece to be welded. In the welding mode depth or keyhole is known to scan the laser beam for a plurality of repetitions in a predetermined welding path, so as to generate progressively the excavation and the molten material. 20 The depth or keyhole welding method is widely used today in many industrial sectors. For example, in the automotive industry it is known the use of infrared laser sources, particularly the near infrared type infrared or NIR (from the English Near Infrared), typically with lengths wavelengths between 700 and 1,400 nm, particularly between 900 nm and 1,080 nm. 25 In the automotive sector, NIR laser welding applications for the modes of surface and / or keyhole welding of metal artifacts include specific engine components, mechanical transmission components, fuel injectors, fuel filters, HVAC components such as air conditioning or structural parts of vehicle frames and bodywork. Typically, the artifacts described above are made, at least in large part, 5 in ferrous alloy materials, such as steel and cast iron. In fact, the absorbance of light infrared radiation generated by NIR sources by iron is relatively high, (about 36.5% for the element iron – Fe), resulting particularly it is convenient to use this technology, which is widely consolidated and with a excellent value for money and high productivity. 10 However, with regard to the technological evolution taking place in the sector automotive, in particular the transition to electric mobility, the market is increasingly introducing a series of new components for electrified propulsion, which require welding processes during the respective assembly phases. A significant portion of these components 15 new generation is made of non-ferrous metal materials. For example, reference is made to electromechanical motors for traction vehicle and electrochemical energy storage devices, such as batteries lithium-ion. In both cases, the need to transmit electrical energy high efficiency requires the use of metallic, non-ferrous materials, 20 high electrical conductivity (low electrical resistivity) in order to reduce transmission losses, essentially due to the Joule effect. Examples concrete examples of these metallic materials are copper (Cu) and aluminium (Al) and the related alloys, which have low electrical resistivity and are economically convenient compared to other materials. 25 However, such non-ferrous materials, such as copper or aluminium, are characterised from the fact that it presents a low absorbance to irradiation in the spectrum of the near infrared (NIR). In particular, when copper is irradiated by a NIR laser source, the absorbance is around 3-4%, in contrast to of the above-mentioned ferrous materials, typically with an absorbance of a orders of magnitude higher than non-ferrous metals such as copper or aluminum. 5 This property inherent to the above non-ferrous materials leads to, of consequently, the need to oversize the NIR laser source single, so as to provide high power and energy densities, so as to combine the low absorbency of the material to be welded. For example, regarding the soldering of the terminals of the elements 10 “hairpin” of the stator windings of a traction motor, the use of is known a NIR laser source. These terminals are made of copper, so given the low absorbance of copper in infrared spectrum the laser welding process is inherently unstable and, consequently, prone to the generation of “spatter” phenomena due to the implosion of molten material in contact with the 15 cloud of vaporized metal during the solidification of the welding bubble. Alternatively, it is possible to use sources operating at wavelengths where the absorbance by the aforementioned non-ferrous materials is greater. Referring to figure 3 of this document, it can be observed that the absorbance of light in non-ferrous metal materials is greater in 20 visible light spectrum (VIS). In other words, for wavelengths between 380 and Around 750 nm, with particular reference to blue light, between 400 and 500 nm wavelength. Advances in the semiconductor-based laser industry have seen since the early 2000s the technological development of Direct Diode Lasers (DDLs), 25 which are applicable to mechanical processes using laser sources and are applicable in various embodiments, such as the blue laser diode or the laser diode green. This type of solution based on laser diodes can be applied on its own, or in combination with an infrared (NIR) laser source 5 above mentioned, to obtain a hybrid welding beam that combines the advantages of both technologies. In such cases two sources are used different, for example a first source operating in the near-wave spectrum infrared (NIR, Near Infrared) and a second source operating in the visible spectrum, especially with blue laser diode light source. 10 On the one hand, the laser diode source operating in the blue light spectrum is mainly used to preheat the artifact to be worked on, stabilizing the process, thanks to the high absorption rate, while the laser source operating in the near-infrared (NIR) provides a higher power high, typically with a diameter of incidence smaller than that of the 15 blue source, which ensures fast and precise melting of the metal for achieve an intrinsically more stable welding process. Furthermore, the presence of a blue laser beam once the metal melting has occurred can contribute to a more controlled cooling process, reducing the porosity of the welding joint and consequently increasing the resistance 20 mechanics of this joint. It is necessary to remember that, whatever the chosen configuration, it is with a single NIR source, or with two sources operating at wavelengths different from each other, the use of laser sources, with particular reference to the way keyhole soldering, the soldering station requires a very fine setting 25 of the welding process. Consequently, it is normally required in parallel a process of control of the welding parameters, so as to be able to react to the various disturbing factors responsible for defects in the manufactured product or soldering station malfunction. In fact, hybrid welding systems, comprising two laser sources operating at different wavelengths, require control of 5 process that guarantees high repeatability and minimization of non- compliance and, at the same time, which reduces energy consumption due to power high NIR lasers and which increase the speed of the process. In particular, this welding process control can be based on optical technologies that collect information from the welding process. 10 For this purpose, it is possible to monitor process signals in welding systems hybrid comprising two laser sources. For example, the patent application US20230311243A1 (see figure 1 of this document) teaches the use of a process sensor in a soldering station comprising two sources operating at different wavelengths, each of which 15 such sources generate a respective laser beam. Each of the two laser beams is defined by several operating parameters, including the beam geometry, more specifically: the diameter incident on the artefact or the focal distance, the pico power and the geometric distribution of the beam (e.g. distribution Gaussian). etc. In US20230311243A1, the two laser beams are combined 20 together in an optical head, typically superimposed concentrically between them. Such optical head is configured to combine both beams and to generate a welding laser beam, incident on the particular artifact to be made the appropriate processes. Once the laser beam acts on the particular artifact to be welded, such 25 artifact reflects a fraction of the incident light. In US20230311243A1, the reflected light is intercepted by an optical sensor, such as a camera or tomographic sensor, configured to detect the intensity of the optical signal in a specific spectrum of wavelengths. US20230311243A1 is particular reference to the use of a camera, in a particular infrared spectrum, to measure the temperature of the molten metal during the welding process. 5 With the configuration just exposed, an intensity verification operation of the detected signal, will concern the whole of the reflected light, allowing solely to identify detectable anomalies in a given spectrum of the reflected light rather than monitoring simultaneously, individually, the correct performance of both laser sources generating the concentric beam 10 of welding. There is therefore a need to perfect a laser welding station that can overcome the drawbacks of the prior art. To do this it is It is necessary to solve the technical problem of building a soldering station laser that is effective, fast and efficient in its operation, with 15 particular reference to a monitoring device configured to detect the light emitted by laser devices and the light reflected by the artefact being processed processing. The main task of this invention is to solve the above problems exhibited, proposing a laser welding station and a related 20 laser welding process that allows the detection of the components of the back-reflected signal in the wavelengths of the two laser sources accidents on the product to be welded. Another object of the present invention is to provide a station for laser welding and a related laser welding process comprising 25 two laser sources operating at different wavelengths that provide richer process information and ability to pinpoint the causes of defects in a more direct way. Another object of the present invention is to provide a station for laser welding and a related laser welding process that provides 5 process information relating to the different phases of the welding process laser. Another object of the present invention is to provide a station for laser welding and a related laser welding process that allows for monitor the fluctuations of the components of the back-reflected signals in the 10 wave frequencies of the main beam and auxiliary beam. Another object of the present invention is to provide a station for laser welding and a related multipurpose laser welding process, which allows you to monitor defects and anomalies in welding processes comprising two laser sources of different wavelengths for different modes 15 of welding and materials to be welded together. A further aim of this invention is to propose a station laser welding and a related laser welding process that guarantees the high productivity of the plant in which it is applied thanks to the drastic reduction of waste or maintenance required due to effects 20 unwanted effects such as spatter. Not least, the aim of this invention is to increase the speed of the process welding using a second laser source appropriately monitored. This task and these goals are achieved by the laser welding station according to claim 1. This task and these purposes are also achieved by means of a laser welding process according to claim 13. To overcome the drawbacks of the known art and to obtain these and 5 further purposes and advantages, the Applicant has studied, tested and realized the present finding. EXHIBITION OF THE FINDING The present invention is expressed and characterized in the claims independent. Dependent claims exhibit other characteristics of the 10 present finding or variations of the main solution idea. In accordance with the above purposes and to solve the above technical problem in new and original way, also obtaining notable advantages compared to the state of the prior art, according to the present invention a station is provided laser welding comprising a first laser device, which comprises 15 a respective oscillator configured to emit a main beam in a first wavelength; a second laser device, which comprises a respective oscillator configured to emit an auxiliary beam in a second wavelength different from the first wavelength; a head optics configured to emit a laser beam to create a well of 20 molten metal to join together at least two elements of an artifact, in which the laser beam is obtained by combining the main beam and the auxiliary; first and second optical organs are configured to transmit the respective first and second beams to the optical head; and a device monitoring configured to detect a reflected light emission from the 25 product in progress. In accordance with an aspect of the present invention, the monitoring device comprises a first semiconductor sensor, configured to detect light reflected in a first spectrum of wavelengths substantially equal to the first wavelength of the main beam, thus monitoring the absorbance 5 of the main beam from the artifact, and a second sensor at semiconductor, configured to detect reflected light in a spectrum of wavelengths substantially equal to the second wavelength of the auxiliary beam, thus monitoring the absorbance of the auxiliary beam by the artifact. 10 Preferably, the first laser device is configured to emit the beam main in the near-infrared (NIR) spectrum. For the near-infrared spectrum infrared refers to a wavelength range between 700 and 1400 nm, preferably between 800 and 1100 nm, even more preferably between 1020 and 1090 nm. 15 Preferably, the oscillator of the first laser device is of the solid-state type. In particular, the oscillator of the first laser device comprises at least one laser diode. Alternatively, the oscillator of the first laser device can include another type of laser element; such as a source of the Nd:YAG type, a fiber laser source or a disk laser. 20 Preferably, the oscillator of the first laser device is configured to operate in single mode. Preferably, the average power density of the main beam emitted by the first laser device is in the range between 15 and 80 MW / cm , in particular between 20 and 2 2 60 MW / cm even more preferably between 25 and 50 MW / cm . For medium density 25 power means the quotient between the average power of the beam and the area on which it affects. To calculate the average power, the integral or summation is performed of the distribution function based on the diameter, i.e. the profile above intensity, as reported in figures 8A, 8B. Preferably, the second laser device is configured to emit the beam auxiliary in the visible light spectrum (VIS). In particular, the spectrum of 5 visible light is that of blue light. Blue light refers to a beam with wavelengths in the range between 400 and 500 nm, even more preferably between 420 and 460 nm. Preferably, the oscillator of the second laser device comprises at least a blue laser diode. Advantageously, the oscillator of the second device 10 laser may comprise a plurality of blue laser diodes. In embodiments wherein the at least one blue laser diode is a plurality, the different laser beams emitted, with limited power, they are combined into a single auxiliary beam thanks to dedicated optical elements, thus obtaining an auxiliary beam of sufficient power for the intended purposes and maintaining acceptable beam quality. 15 Preferably, the average power density of the auxiliary beam emitted by the second laser device is in the range between 0.5 and 1.5 MW / cm , even more preferably between 0.6 and 1 MW / cm . Preferably, the first laser device comprises, downstream of the respective oscillator, a first optical group. 20 Preferably, the second laser device comprises, downstream of the respective oscillator, a second optical group. Preferably, the first and / or second optical group comprises a fiber optic cable optics. This fiber optic cable is configured to transmit the laser beam generated by the respective oscillator. 25 Preferably, the first optical group comprises a first collimating lens. Preferably, the second optical group includes a second lens collimator. The collimating lenses of the respective optical groups are configured to straighten the beams coming from the respective oscillators of the laser devices. 5 Advantageously, the first and / or second light cluster may comprise a respective convex lens. This convex lens is configured to make converge the beam to a predetermined focal distance. Some forms implementations provide that this convex lens is mounted movably, so to be able to vary the focal length of the beam of the respective laser device. 10 Preferably, the first laser device is configured to emit a beam main laser having a substantially circular area. Preferably, the second laser device is configured to emit a auxiliary laser beam having a substantially circular area. Preferably, the diameter of the main beam emitted by the first device 15 laser will have a smaller diameter than the auxiliary beam diameter emitted by the second laser device. Preferably, the first laser device is configured to emit a beam principal of the Gaussian type. A Gaussian light beam is the beam having an intensity profile perpendicular to the direction of 20 beam propagation with Gaussian distribution (Gaussian bell). Preferably, the second laser device is configured to emit a Gaussian type auxiliary beam. Advantageously, the first and / or second optical group may comprise, a valley of the respective focal lens, a diffractive optical element (in English Diffractive Optical Element, DOE). This diffractive optical element or DOE is configured to shape the profile of the main and / or auxiliary beam. The use of a diffractive optical element or DOE is particularly advantageous in ring-core configurations, in which they are 5 two concentric beams of different diameters are superimposed on each other. In such cases, it is it is possible to modify the propagation profile of the external beam so as to create an outer ring, in which the beam distribution in the direction perpendicular to the propagation plane has a W shape, leaving the substantially empty central part. In such cases, the smallest diameter bundle 10 can occupy the inner (empty) part of the ring and act as a nucleus. Preferably, the first and / or second light clusters comprise respective mirrors. The mirrors of the respective first and second devices are configured to direct the respective main beam and auxiliary beam towards the optical head of the soldering station. 15 Preferably, such first and / or second optical group mirrors are configured to orient the main beam and the auxiliary beam so coaxial. Preferably, the welding head comprises at least one welding lens configured to direct the laser beam towards the workpiece to be machined. 20 Preferably, the elements of the artifact to be welded together will be positioned close to each other. According to preferred embodiments of the present invention, the elements of the artifact to be welded together are both arranged facing the welding head, oriented so as to be hit directly by the beam laser. These embodiments allow the processing to be carried out in a of thermal type welding or in depth or keyhole welding mode. Alternative embodiments of the present invention provide that the elements of the artifact to be welded together are arranged facing the head 5 of welding and stacked on top of each other. This way at least a first element will be placed between the welding head and a second element, on where the first element will lie. In these embodiments, the welding beam will invest the first element, the heat transfer by conduction between the first and the second element of the artifact downstream of the incidence of the beam of 10 welding on the first element generating a superficial molten metal well will determine the mutual welding between the two elements. Preferably, the at least one welding lens of the welding head is constituted as a plurality of lenses. In particular, such plurality of lenses will be of the f-theta type. f-theta lenses contain a plurality of lenses, typically three 15 or four, pre-aligned with each other and housed in a sturdy case and are configured to focus the welding beam onto a flat surface. This This type of configuration is particularly advantageous when the artifact to be work is arranged on a plane perpendicular to the welding beam at a predetermined distance. 20 Preferably, the welding head comprises a scanning device upstream of the welding lens. This scanning device is configured to change the orientation of the welding beam, i.e. the main and auxiliary combined together in the welding head, so as to scan the welding beam along a path on the surface of the artifact 25 work. Preferably, the scanning device is a galvo type scanner. Such galvo type scanner is configured to vary the orientation of the beam of welding in two directions perpendicular to each other, so as to scan the welding beam on two mutually perpendicular directions of the surface of the 5 artifact to be worked on. In accordance with another aspect of the present invention, the device of Soldering station monitoring also includes a mirror monitoring configured to reflect the light of the main beam and the beam auxiliary and the light reflected from said component towards at least said first and 10 said second semiconductor sensors of the monitoring device. In This way, the back-reflection signals will be efficiently transmitted to the individual semiconductor sensors of the monitoring device. Preferably, the tracking mirror is aligned with the mirrors of the respective first and second optical groups, so as to allow both to be reflected 15 the light of the main and auxiliary laser beams, both the light reflected from the artifact and crossing the welding head. This monitoring mirror can be placed between the laser devices and the welding head. Alternatively, such a monitoring mirror may be placed between the two mirrors of the respective first and second optical groups 20 of the laser devices. In accordance with another aspect of the present invention, the device of monitoring also includes a first optical filter placed between the mirror monitoring and the first and second semiconductor sensor. This first optical filter is configured to reflect at least a second back-reflecting signal 25 reflection towards the second semiconductor sensor and to transmit at least a first retro-reflection signal towards the first sensor semiconductor. By doing so, the first optical filter separates the light into two spectra complementary, facilitating the separate detection of signals. Preferably, the first optical filter is a dichroic filter configured for function as a low-pass filter (from the English, low band pass filter) beyond 5 below a certain threshold wavelength, selectively reflecting light with wavelengths lower / higher than the wavelength of that threshold and to transmit light with wavelengths higher / lower than that theshold. Alternative embodiments provide that such first optical filter comprises 10 a beam splitter. This beam splitter will be configured to split the incident light on the same into two or more beams, each beam having a certain spectrum of wavelengths. In the embodiments in where beam splitters are used, it will also be necessary, the use of additional optical filters placed downstream of the beam splitters. In 15 alternative, other forms of implementation other than the filter are not excluded dichroic or those including a beam splitter just illustrated. In accordance with another aspect of the present invention, the device of monitoring also includes a third semiconductor sensor, configured to detect the light emitted by the welding process in a third 20 visible light spectrum, so as to monitor the plasma of the metal well fused generated by the incidence of the main beam and the auxiliary beam on the artifact. This third visible light spectrum is an indicator of plasma generated by the incidence of the welding beam on the product to be worked during fusion. Typically, the plasma is detectable in the light spectrum 25 visible. The detection of light in the third visible light spectrum from the third semiconductor sensor will allow to monitor the welding process for as regards the plasma of the molten metal well generated by the incidence of the main beam and the auxiliary beam, i.e. the combined welding beam. In accordance with another aspect of the present invention, the device of monitoring also includes a second optical filter placed between the first 5 optical filter and the second semiconductor sensor. This second optical filter is configured to reflect the second emission back-reflected signal of light with a spectrum of wavelengths substantially coinciding with the second wavelength emitted by the second laser device. In accordance with another aspect of the present invention, the second optical filter is 10 further configured to transmit a third light plasma signal emitted in the third spectrum of visible light with longer wavelengths compared to those of the second spectrum. This second optical filter is arranged in so as to orient the third plasma signal towards the third sensor semiconductor (55). 15 In accordance with another aspect of the present invention, the device of monitoring also includes a fourth semiconductor sensor configured to detect the light emitted by the welding process in a quarter thermal infrared spectrum in order to monitor the temperature of the artifact. Monitoring the temperature of the product concerns the different phases of the 20 laser welding process, measuring the temperature of the metal well melt formed on the artifact and the temperature of the soldering joint during the subsequent cooling once the main and / or auxiliary beams have been switched off. In accordance with another aspect of the present invention, the device of monitoring further includes a third optical filter placed between the 25 first optical filter and the first semiconductor sensor. This third optical filter is configured to reflect the first back-reflected signal of the emission of reflected light with a substantially coincident wavelength spectrum with the first wavelength emitted by the first laser device. In accordance with another aspect of the present invention, the third optical filter is further configured to transmit a fourth thermal infrared signal 5 emitted by the welding process in the fourth thermal infrared spectrum with wavelengths longer than those of the first light spectrum, this third optical filter is arranged to orient the fourth thermal infrared signal towards the fourth semiconductor sensor. Preferably, the first semiconductor sensor comprises a photodiode. 10 Preferably, the second semiconductor sensor comprises a photodiode. Preferably, the photodiodes of the first and second semiconductor sensors they present a first band of responsiveness that is substantially equal to each other. Preferably, the third wire sensor is a photodiode. 15 Preferably, the photodiode of the third semiconductor sensor has a responsiveness band substantially equal to the first responsiveness band. Preferably, the fourth wire sensor is a photodiode. Preferably, the photodiode of the fourth semiconductor sensor has a second band of responsiveness in a different wavelength spectrum than 20 first response band of the first and second semiconductor sensors. Preferably, the soldering station includes a controller. Preferably, the controller is operationally connected to the device monitoring. Preferably, the controller is operationally connected to the first and / or second laser device. Preferably, the controller is operationally connected to an interface man-machine. 5 Preferably, the controller includes a memory for storing data and programs and a processor to perform computational operations. Preferably, the controller is configured to record schedules predetermined welding. These predetermined welding programs contain the operating parameters, timing, etc. of the first and 10 second laser devices involved in the welding process as a function of the artifact to be worked. Preferably, the controller is configured to calculate a first band of tolerance of the signal recorded by the first semiconductor sensor based on to the operating parameters of the first laser device of the relevant 15 welding program and the reflectance of the material on which the beam strikes main in the first wavelength. Preferably, the controller is configured to calculate a second signal tolerance band recorded by the first semiconductor sensor based on the parameters of operation of the second laser device of the relevant program 20 welding and the reflectance of the material on which the auxiliary beam strikes in the second wavelength. By doing so, it will be possible to verify anomalies, or values ​​outside the threshold, in the welding process through signal analysis detected by the first and second semiconductor sensors, in the absence of a phase of training or comparisons of real signals coming from the field with 25 complex reference signals. Preferably, the controller is configured to record actual detected signals from semiconductor sensors. Each of these real detected signals is a time trend consisting of one of two things: a continuous time series and a plurality of consecutive discrete values ​​of a discrete time series. 5 Preferably, the controller is configured to record a first signal and a second real signal relating to a first wave longitude and a second wave longitude, essentially corresponding to the longitudes waveform of the main beam and auxiliary beam of the welding beam. Preferably, the controller is configured to additionally record a third 10 real plasma signal in a third of the visible light spectrum. Preferably, the controller is configured to additionally record a fourth real thermal signal in a fourth spectrum of thermal infrared light. Preferably, the controller contains reference signals in memory predetermined welding programs. Each of these reference signals 15 is a time trend consisting of one of a continuous time series and a plurality of consecutive discrete values ​​of a discrete time series referring to a welding program. Preferably, the controller is configured to store a first signal and a second signal reference relating to first light spectrum and a second light spectrum, 20 substantially corresponding to the wavelengths of the main beam and auxiliary beam of the welding beam of a predetermined program of welding. Preferably, the controller is configured to additionally record a third plasma reference signal. Preferably, the controller is configured to additionally record a fourth thermal reference signal. 25 Preferably, the controller is configured to split the real signals and the reference signals in a plurality of partial segments, as a function of instants predefined according to the predetermined welding programs, designed to identify anomalies and defects in the time intervals in which they can occur. Preferably, the controller is configured to extract features from the real signals and reference signals. These characteristics can be 5 referring to the entire time series, discrete or continuous, or in a segment predetermined partial. Preferably, the controller is configured to compare characteristics of at least one real signal with respective characteristics of a signal reference. The result of this comparison will allow us to verify whether the 10 whether the detected signal is as expected or whether there is an anomaly in progress. The present invention extends its scope of protection to a welding method laser for welding together at least two elements of an artifact, comprising the phases of: placing at least two elements of an artefact close together in an area 15 welding and facing an optical head of a laser welding system; scan at least one of the at least two elements of the artifact with a laser welding beam to create a well of molten metal on itself; cool the molten metal well by turning off the welding laser beam to obtain a solidified solder joint; and monitor a light emission 20 reflected during the scanning and cooling phases to evaluate the quality of called solder joint. The welding laser beam consists of at least one main beam, emitted from a first laser device at a first wavelength, and a beam auxiliary, emitted by a second laser device at a second length 25 wavelengths. Each of the main and auxiliary beams will have a set of predetermined operating parameters, including average power density, beam distribution, focal length, etc. Preferably, the first wavelength belongs to the light spectrum in the near infrared (NIR). 5 Preferably, the second wavelength belongs to the visible spectrum (VIS). According to a particularly interesting embodiment, the second wavelength belongs to the blue light spectrum. It means for the light spectrum of blue light to light with a wavelength between 400 and 500 nm, more specifically between 420 and 460 nm. 10 In accordance with one aspect of the present invention, the monitoring phase is composed of the sub-phases of - monitor an initial signal of retro-reflection of the light emission reflected, where said first retro-reflected signal is formed by an electromagnetic emission with a wavelength substantially 15 coinciding with the first wavelength emitted by the said first laser device; and - monitor a second back-reflected signal of the emission reflected light, wherein said second back-reflected signal is formed from an electromagnetic emission with wavelength 20 substantially coinciding with the second wavelength emitted from said second laser device. In accordance with one aspect of the present invention, the sub-phase of monitoring a first signal of retro-reflection of the reflected light emission consists of the calculate a first tolerance band of the signal recorded by the first 25 semiconductor sensor based on the operating parameters of the first laser device of the relative scanning phase and the reflectance of the material on which the main beam in the first wavelength is incident. In accordance with one aspect of the present invention, the sub-phase of monitoring a second retro-reflection signal of the reflected light emission consists of the 5 calculate a second tolerance band of the signal recorded by the second semiconductor sensor based on the operating parameters of the second laser device of the relevant welding program and the reflectance of the material on which the auxiliary beam in the second wavelength is incident. By doing so it is possible to detect the presence of an anomaly during the phase 10 to monitor the laser beam, if the first and / or second back-up signal reflection goes outside the pre-established acceptance band. In this way, there will be no need to train a model based on a set historical data or using a reference signal to compare signals to of back-reflection recorded by semiconductor sensors. 15 Preferably, the sub-phase of monitoring a first retro-reflection signal and the sub-phase of monitoring a second retro-reflection signal occurs at the same time. In accordance with one aspect of the present invention, the monitoring phase It also includes the sub-phase of monitoring a third plasma signal of the 20 emitted light (EL) due to the scanning phase, where this third signal plasma is formed by electromagnetic emission in the light spectrum visible. Preferably, the plasma signal is an electromagnetic emission in a third spectrum of light with wavelengths shorter than the first wavelength of the first retro-reflected signal and longer than at the second wavelength of the second back-reflected signal. In accordance with one aspect of the present invention, the monitoring phase It also includes the sub-phase of monitoring a fourth infrared signal 5 thermal light emitted by the welding process, in which the infrared signal thermal is formed by an electromagnetic emission in the infrared spectrum and different from the first wavelength. Preferably, the thermal infrared signal is an electromagnetic emission in a fourth spectrum of light with wavelengths longer than the 10 first wavelength of the first back-reflected signal. In accordance with an aspect of the present invention, each of the sub-phases of monitoring includes the steps of: detecting and acquiring signals coming from the artifact for the duration of the scanning and cooling phases of each welding operation between the elements of the artefacts to be welded 15 mutually; obtain the real signals relating to the respective light spectra from the signals reflected and emitted by the artifact, where the real signals are a time trend consisting of one of two things: a continuous time series and a plurality of consecutive discrete values ​​of a discrete time series; compare characteristics of at least one time trend of real signals 20 with respective characteristics of at least one respective temporal trend of the reference signals, identifying defects in the welding process artifact based on the relationship between the compared characteristics; and classify the real signals based on the comparison of such characteristics of the real signals with the respective characteristics of the reference signals, identifying any defects 25 and / or estimating mechanical properties of the weld of the artifact based on the unique relationship between the compared characteristics. In accordance with an aspect of the present invention, the laser welding method further includes, before the scanning and cooling phases the phase to select a predefined welding program, where that program welding includes instructions and operating parameters with which a 5 controller performs the scanning and cooling phases, where such program welding further comprises at least one reference signal of the reflected signal related to the predetermined welding program, and where such welding program further includes at least one window temporal ti-tf monitoring in such steps of scanning and cooling, such 10 ti-tf time window corresponding to a stage of the welding operation with initial time and final time. In accordance with one aspect of the present invention, the monitoring phase further includes, before the step of comparing, the steps of: extracting values ​​relating to the stage defined by at least one ti-tf time window 15 of the at least one real signal and the at least one respective reference signal of each welding operation; and calculate at least one physical indicator relating to the stage of the welding operation, starting from the values ​​enclosed in the at least one ti-tf time window and extracted from the at least one real signal and from at least one respective reference signal. 20 In accordance with one aspect of the present invention, the step of comparing further includes the step of comparing the at least one physical indicator of the stage of the at least one real signal with the at least one respective indicator physical of the same stage as the at least one reference signal. According to embodiments of the present invention, the method provides a phase 25 training sessions using real welding process data made with the first laser device and the second laser device. In this training phase the method includes a campaign to collect the signals, in particular of the first retro-reflection signal, of the second retro-reflection signal, third plasma signal and fourth signal thermal infrared. 5 This data collection campaign will be carried out for a programme of predetermined welding, that is, the set of operating parameters for the creation of a weld of a particular artifact. The campaign of data collection will include both data on acceptable welding cycles, and they will be labelled as such, and data relating to welding cycles with defects present 10 or anomalies, which will also be labeled as such. The combination of data from the first retro-reflection signal, to the second retro-reflection signal, third plasma signal and fourth thermal infrared signal will allow to identify defects and / or anomalies associate these defects and / or anomalies with their causes. Starting from data analysis 15 historical ones labeled as acceptable and anomalous / defective will it be possible to build a model to identify such anomalies / defects. Preferably, a machine learning approach will be followed supervised, or Machine Learning. This Machine Learning model will be compared with the actual signals coming from the welding process. 20 Preferably, the method involves the use of data from at least one of the signals (first, second, third and / or fourth) collected in the Machine Learning model. In some embodiments, the signal (first, second, third and / or fourth) of the entire welding process cycle. In other alternative embodiments, partial data of at least 25 a signal (first, second, third and / or fourth), i.e. referring only to one interval predetermined temporary interval of the entire cycle. For temporary interval means, for example, each of the welding stages, such as the pre-melting heating stage, the welding stage (molten metal), the joint cooling stage, the temporary interval of long passage 5 the gap between two elements of the product to be welded, etc. Preferably, a machine learning approach will be followed supervised, or Machine Learning. This Machine Learning approach provides for the use of at least one algorithm for the detection of an anomaly / defect. 10 Some embodiments provide that the training phase includes the use of an unsupervised approach. This approach will be based on data of the four signals, in particular the first retro-reflection signal, the second retro-reflection signal, the third plasma signal and the fourth thermal infrared signal. The data relating to the four signals will be used, 15 to define, for each type of anomaly / defect, the acceptability bands. These acceptance bands will be based on the physical behavior of the four signals (physics-informed approach) for the welding program predetermined. In fact, such acceptance bands can be enclosed within a specific time interval of a welding cycle, doing 20 reference to one or more stages of the welding process (heating, fusion, cooling). The output of the signal values ​​(first, second, third and / or fourth) with respect to the acceptability bands will indicate the presence of an anomaly / defect starting from the modeling of such anomaly / defect based on to the acceptance bands. 2 9 ILLUSTRA TION OF SIGNS F urther charac teristics which eva nt additions of the find will be added to the next year e vi de nti all aluced ell ades cri zi oned ett a gli at a di unaf or ma di re ali zza zi one pr ef erit amanones cl u si va di un pr oce di me nt op er las al d at ur alas er secondo 5 il tr ov at o, ill u str at oa tit ol o di esem pi onon li mit ati voconl' au sili od ell eu nit e tav ol e di di segnoinc ui:  la fi g. 1 rap pr ese nt a, in vi st aschem ati caun 'app ar ecc hi at ur a di s al d at ur alas er pr ese nt en ell' art en ot a ;  la fi g. 2 rap pr ese nt a, in vi st aisom etri ca, unarap pr ese nt a zi one 10 scheme ati ca di un pr ocesso di s al d at ur alas er inmodo di pr of on dit à okeyh ol e ;  la fi g. 3 rap pr ese nt a gr afi came nt elaass or banzad ell alucein ci de nt ein function of the wavelength that eluces a plurality of materials metal allicin in the visible spectrum (VI S) and infrared;  the daughter g. 4 schematic showing the welding station 1 5 according to the present finding ;  the daughter g. 5 represents optical schematics of a primary the laser is configured to emphasize the main laser.  the daughter g. 6 The rap presents the same as the eighth one-second pop up to configure the auxiliary bundles as follows. 2 0 preferred active form of the present finding.  the daughter g. 6 B rap presents as chemically as the eighth one-second pop up to configure the auxiliary bundles as follows. the present finding was an alternative assembly. 3 0  the daughter g. 7 represents schematically the bundle or the main bundle. auxiliary air emitted from the welding station presently found according to f or more preferred sea.  the daughter g. 8 A graphical representation of power density distribution 5 of the main bands for the silicon water to follow a preferred tidal pattern at the present and found.  the daughter g. 8 Graphical presentation of power density distribution of the main bands for the silicon water to follow an alternating activated tide present and found. 1 0  the fi g. 9 schematically shows the alignment solder bundle incident on the elements immediately adjacent to a man's act s al d at ur ainmodo pr of ondookeyh ol e .  the daughter g. 1 0 shows schematically the solder bundle at the bottom incidence of a first element placed between the beam and a second element 1 5 of the manufacturer, for the thermal ainmode dator.  l a fi g. 1 1 r a p pr e s e nt a s c h e m ati c a m e nt e u n d ett a gli o d el di s p o siti v o di m o nit or a g gi o d ell a st a zi o n e di s al d at ur a l a s er r a p pr e s e nt at o i n fi g ur a 1 2 .  l a fi g. 1 2 r a p pr e s e nt a s c h e m ati c a m e nt e l a st a zi o n e di s al d at ur a l a s er s e c o n d o u n a f or m a r e ali z z ati v a pr ef erit a d el pr e s e nt e tr o v at o; 2 0  l a fi g. 1 3 r a p pr e s e nt a s c h e m ati c a m e nt e l a st a zi o n e di s al d at ur a l a s er s e c o n d o u n a f or m a r e ali z z ati v a d el pr e s e nt e tr o v at o;  l a fi g. 1 4 r a p pr e s e nt a s c h e m ati c a m e nt e l a st a zi o n e di s al d at ur a l a s er s e c o n d o u n a f or m a r e ali z z ati v a pr ef erit a d el pr e s e nt e tr o v at o; 3 1  the daughter g. 1 5 shows the schematics of the device model of the welding station clock is shown in Figure  The daughter. 1 6 represents the components of the respective first signal back - reflection, back-reflection second signal, plasma tertiary signal 5 fourth sign on the infrared spectrum related to the operation performed in the first second of spo sites fly er.  the daughter g. 1 7 graphical representation of signals transmitted to sensors if I am driving and the monitoring device.  the daughter. 1 8 A , 1 8 B , 1 8 C , and 1 8 D graphically present various signals 1 0 transmitted to the conductive sensors, threshold wavelength at el respective eight-sky filter and minimum maximum signal responsiveness of the respective conductive sensors.  the daughter. 1 9 A, 1 9 B and 1 9 Figs show graphically components of the first back-reflection signal (at the first wavelength λ) relative to 1 5 main bundle to the first flying device and to the auxiliary bundle emitted at the second of the flying sites, so as the sum of the entrances compose you.  the daughter. 2 0 A, 2 0 B and 2 0 C graphically represent components of the back-reflection second signal (at the second wavelength λ 2 ) 2 0 relative to the main beam and emitted from the first flying device in the beam the sili are emitted to the second of the volatile sites, which is the sum of and ntr composed nti.  the daughter. 2 1 A, 2 1 B and 2 1 Figures graphically show components of the plasma signal (in the spectrum of the visible light WB 3 ) relative 2 5 main bundle to the first flying device and to the auxiliary bundle 3 2 emitted at the second of the flying sites, so as the sum of the entrances compose you.  the daughter. 2 2 A, 2 2 B and 2 2 Figures show graphically components of the fourth infrared signal (unknown infrared spectrum 5 WB 4 ) relative to the main beam when the first device is flown to the beam the siliques are emitted to the second of the volatile sites, which is the sum of and ntr composed nti.  the daughter g. 2 3 represents the first detected back-reflection area signal from the first sensor or ease of conduction and the second signal to the rear air - 1 0 reflection detected in the second sense or ease of conduction in different setups of the welding process.  the fig. 2 4 represents the first and second signals of back -reflections, the value of hearing and signaling expectations as well as acceptance ability makes it possible to hide two signals. 1 5  the fi g. 2 5 represents the first-second back-reflection signals, and a series of physical indicators related to specific stages of the welding process.  the daughter g. 2 6 represents the tertiary signal of plasma, including elastic real time, reference time series and relative physical indicators the welding process is created. 2 0  the fi g. 2 7 represents the fourth sign in the oxothermal infrastructure, regardless of real time laser, reference time laser and physical indicators related to the welding process.  the daughter. 28 and 2 9 represent the first and second signals of retro- reflections, the respective reference signals and the acceptance band 2 5 can be used to hide the two reference signals. 3 3  the daughter g. 3 0 reports the tabulation of possible cone-generated defects. welding station according to the present findings and the respective signals for detect such defects with any known monitoring system according to the present invention. 5 CURRENT REA FORM DESCRIPTION FOUND Figure 1 shows a welding station at the laser of the artenote. P on simplicity in the description of such a station a symbol will be used analogous to that of the welding station according to the present invention. L a 1 0 welding station 1 of the arteno comprises a first flying device 2, configure to send a first phase to B 1 as a first error of wave λ , each second of flying devices er 3, configured to emit a second phase there waves er B 1 onesecond wavelength λ T hey were free second phases B 1, B 2 sonocombined welding station 1 5 configure after plugging in the solder wire to the component lav or ar e. Welding station 1 further comprises a device of Clock monitor 5 is configured to detect the light reflected from the component a predicted wavelength spectrum. T al econ figures the observed allows single monitoring of reflected light components relative to the 2 0 first phase to the second phase. Referring to Figure 2, the elements are shown schematically. representative of a soldering operation while the driver is unmodeled to the pr of on said to okeyh ol e , inc ui unfas ci olas er LB pen etr ain pr of on said to inun composes the work, digging outside the narrow, called okeyhole, 2 5 while at the same time generating, preventing the interaction of LB e component W a metal evaporation, unionization of the surrounding environment in the form of plasmas and a well of molten metal. He parts already subjected to the incidence of the LB laser beam on the W component si they cool fast originating a soldering joint. With reference to figure 3, the absorbances are represented graphically 5 characteristics of certain metallic materials in terms of light emissions as a function of wavelength of the incident light. This graph represents the values recorded in the visible and infrared spectrum. It should be noted that for metals not ferrous and related alloys, the characteristic absorbance to light in the spectrum of the near infrared (NIR) is particularly low. For example, in the case of the 10 copper (Cu) and its alloys, the absorbance in the spectrum between 900 and 1,000 nm is at below 10%. Following the example of copper (Cu), as regards the light in the visible spectrum (VIS) the absorbance value is much higher, ton rates above 60%. For this reason, the recent development of the diode laser blue is gaining traction in combination with NIR laser sources for applications 15 of welding. With reference to figures 4, a preferred embodiment is shown of the welding station 1 according to the present invention. This welding station Welding 1 includes: − a first laser device 2, which comprises a respective oscillator 20 21 configured to emit a main beam B1 in a first wavelength λ1, where the first laser device 2 is preferably configured to emit such a main beam B1 in the near spectrum infrared (NIR); − a second laser device 3 comprising a respective oscillator 25 31 configured to emit an auxiliary beam B2 in a second wavelength λ2 different from the first wavelength λ1, where the second laser device 3 is preferably configured to emit such auxiliary beam B2 in the visible light spectrum (VIS), in particular in the blue light spectrum; − an optical head 4 configured to emit a laser beam LB in a 5 to create a well of molten metal to join together at least two elements W1, W2 of an artifact W, such laser beam LB being obtained by combining together the main beam B1 and the auxiliary beam B2; − first 22 and second 32 optical organs configured to transmit their respective 10 first B1 and second B2 beam to optical head 4; and − a monitoring device 5 configured to detect an emission of reflected light RL from the manufactured product W; wherein such a device Monitoring 5 further includes: − A first semiconductor sensor 52, configured to detect the 15 RL reflected light in a first wavelength spectrum WB1 substantially equal to the first wavelength λ1 of the beam main B1, thus monitoring the absorbance of the main beam B1 by artifact W; and − A second semiconductor sensor 53, configured to detect 20 the reflected light RL in a second wavelength spectrum WB2 substantially equal to the second wavelength λ 2 of the beam auxiliary B2, thus monitoring the absorbance of the auxiliary B2 beam from the W artifact. The first laser device 2 comprises, downstream of the respective oscillator 21, a 25 first optical group 22. The first 22 optical group includes a fiber optic cable 23. The second 3 laser device comprises, downstream of the respective oscillator 31, a second optical group 33. The second 32 optical group includes a fiber optic cable 33. Fiber optic cables 23, 33 are configured to transmit laser beam B1, 5 B2 generated by the respective oscillator 21, 31 with elements arranged downstream of these fiber optic cables 23, 33. The first 22 and / or second 32 light clusters include respective mirrors 24, 34. The mirrors 24, 34 of the respective first 22 and second 32 devices are configured to orient the respective main beam B1 and auxiliary beam B2 10 towards the optical head 4 of the soldering station 1. Such mirrors 24, 34 of the first 22 and / or second 32 optical group are configured to orient the main beam B1 and the auxiliary beam B2 coaxially, as shown in figure 17. The welding head 4 comprises at least one welding lens 41 15 configured to direct the welding laser beam LB towards the artifact W from work. The welding head 4 also comprises a device scanning 42 upstream of the welding lens 41. This scanning device 42 is configured to change the orientation of the welding beam, i.e. the main beam B1 and auxiliary beam B2 combined together in the welding head, in 20 way to scan the LB welding beam along a path on the surface of the artifact W to be worked. With reference to figures 5 and 6A) The first optical group 22 also comprises Optics 25 of the first laser device. Optics 25 of the first laser device include first collimating lenses 26. The second optical group 32 It also includes optics 35 of the second laser device. The optics 35 of the second laser device include second collimating lenses 36. The first 26 and second 36 collimating lenses of the respective first 22 and second 32 light clusters are configured to straighten the respective main beams B1 and 5 secondary beam B2 coming from the respective first 21 and second 31 oscillators of the respective laser devices 2, 3. The first 22 and / or second 32 light clusters may also include a respective convex lens. This convex lens is configured to make converge the beam to a predetermined focal distance. Some forms 10 embodiments provide that this convex lens is mounted movably, so to be able to vary the focal length of the beam of the respective laser device. Referring to figure 7, the first laser device 2 is configured to emit a main laser beam B1 having an area substantially circular. Furthermore, the second laser device 2 is configured to emit a 15 auxiliary laser beam B2 having a substantially circular area. Preferably, the diameter D1 of the main beam B1 emitted by the first laser device 2 will have a smaller diameter than the diameter D2 of the auxiliary beam B2 emitted by the second laser device 3. With reference to figures 8A and 8B, the first laser device 2 is configured 20 to emit a Gaussian type main beam B1. For light beam Gaussian means the beam having a BS1 intensity profile perpendicular to the direction of propagation of the beam B1 with distribution Gaussian (Gaussian bell). Similarly, the second laser device 3 is configured to emit an auxiliary beam B2 of the Gaussian type. In other 25 words, auxiliary beam B2 has a perpendicular BS2 intensity profile to the direction of propagation of the B2 beam with Gaussian distribution (a Gaussian bell). With reference to figures 6B and 8B, the first 22 and / or second 32 light cluster may comprise, in an alternative embodiment, an optical element 5 diffractive (in English Diffractive Optical Element, DOE) placed downstream of the respective lenses. This diffractive optical element or DOE is configured to shape the profile of the main and / or auxiliary beam. In particular, the second optical group 32 may comprise, downstream of the collimating lenses 36, a diffractive optical element 37 configured to modify the BS2 beam profile 10 auxiliary B2. Figure 8B shows such a double W profile, which acts as a ring for a core of the main beam B1, with Gaussian profile BS1. With reference to figures 9 and 10, the elements W1, W2 of the artifact W from to weld together they will be positioned close to each other. According to shapes realization of the present invention, the elements W1, W2 of the artifact W from 15 mutually solderable are both arranged facing the head of welding 4, oriented so as to be directly hit by the laser beam LB (see figure 9). These embodiments allow the realization of the processing in a thermal welding mode or in a welding mode depth or keyhole. 20 Alternative embodiments of the present invention provide that the elements W1, W2 of the W artifact to be welded to each other are arranged facing each other to the welding head and stacked on top of each other. This way at least one first element W1 will be placed between the welding head 4 and a second element W2, on which the first element W1 will lie. In these embodiments, the 25 LB welding beam will hit the first element W1, heat transfer for conduction between the first W1 and the second W2 element of the artifact W a valley of the incidence of the welding beam LB on the first element generating a surface molten metal well will cause the two to weld together elements. The monitoring device 5 of the welding station 1 further comprises: 5 a tracking mirror 51 configured to reflect the beam light main beam B1 and auxiliary beam B2 and the reflected light RL from said W component towards the first 52 and the second 53 semiconductor sensors of the monitoring device 5. The monitoring mirror 51 is aligned with the mirrors 24, 34 of the respective 10 first 22 and second 32 light groups, so as to allow to reflect both the light of the main laser beams B1 and auxiliary B2, both the reflected light RL from the artifact W and passing through the welding head 4. The monitoring mirror 51 may be arranged between the two laser devices 2, 3 and the welding head 4. Alternatively, such monitoring mirror 15 can be placed between the two mirrors 24, 34 of the respective first 22 and second 32 optical groups of laser devices 2, 3. The soldering station also includes a controller 6. This controller is operationally connected to the monitoring device 5. Furthermore, the controller 6 is operationally connected to the first 2 and / or second 3 device 20 lasers. The controller 6 can be operationally connected to an interface man-machine, not represented in the figures. With reference to figures 11 and 12, the monitoring device 5 comprises furthermore a first optical filter 54 interposed between the monitoring mirror 51 and the first 52 and the second semiconductor sensor 53. This first optical filter 25 54 is configured to reflect at least one second back-reflected signal RL2 to the second semiconductor sensor 53 and transmit at least one first retro-reflection signal RL1 towards the first sensor 52 a semiconductor. By doing so, the first optical filter 54 separates the light into two complementary spectra, facilitating the separate detection of RL1, RL2 signals. 5 With reference to figure 13, some embodiments of the invention provide that the monitoring device further comprises a third sensor 55 a semiconductor, configured to detect reflected light EL3 in a third spectrum of visible light visible WB3. This reflected light EL3 in the third light spectrum visible WB3 is an indicator of the plasma generated by the incidence of the beam of 10 LB welding on the W artifact to be worked. Some embodiments of the present invention provide that the device monitoring also includes a fourth semiconductor sensor 56 configured to detect EL4 reflected light in a fourth WB4 spectrum of light infrared thermal in order to monitor the temperature of the artifact W 15 during the welding process, i.e. the temperature of the respective well molten metal formed on it and the temperature during the next cooling once the main beams B1 and auxiliary beams B2 are switched off soldering station 1. The fourth infrared light spectrum WB4 includes a range with wavelengths longer than the wavelengths of the 20 first WB1 spectrum. With reference to figures 14 and 15, the monitoring device 5 comprises furthermore a second optical filter 57 placed between the first optical filter 54 and the second semiconductor sensor 53. This second optical filter 54 is configured to reflect the second back-reflection signal RL2 25 of the reflected light emission RL with a spectrum of wavelengths substantially coinciding with the second wavelength λ2 emitted by the second laser device 3. The second optical filter 57 is advantageously configured to transmit a third EL3 plasma signal with wavelength spectrum longer than 5 that of the first wavelength λ. This third plasma signal EL3 is oriented towards the third semiconductor sensor 55 to detect the third EL3 plasma signal. Similarly, the monitoring device 5 further comprises a third optical filter 58 placed between the first optical filter 54 and the first sensor 10 semiconductor. This third optical filter 58 is configured to reflect the first RL1 back-reflection signal of the reflected light emission RL with one wavelength spectrum substantially coinciding with the first wavelength λ emitted by the first laser device 2. Additionally, the third optical filter 58 is configured to transmit a fourth signal 15 thermal infrared EL4 with a fourth thermal infrared spectrum WB4. This fourth spectrum WB4 refers to wavelengths longer than those of the first WB1 spectrum. The third optical filter 58 is oriented towards the fourth sensor semiconductor 56 for detecting said fourth thermal infrared signal EL4. Referring to figure 13, the reflected light relative to a 20 laser process with two laser sources operating at different wavelengths between them and applied to a manufactured product. By way of example and not limited to limitingly, a particularly interesting embodiment is reported in which the beam main B1 is generated by an oscillator 21 of the first laser device 2 in the infrared spectrum and in which the auxiliary beam B2 is generated by an oscillator 31 25 of the second laser device 3 in the blue light spectrum. In particular, Figure 13 reports the signal intensity of four RL1 channels, RL2, EL3 and EL4 of the reflected light as a function of the incident wavelength on the respective semiconductor sensors 52, 53, 55, 56. The first channel RL1 represents the retro-reflection signal of the main beam light B1 5 (operating in a first wavelength λ1) incident on the first sensor at semiconductor 52. The second channel RL2 represents the feedback signal reflection of the auxiliary beam light (working in a second wavelength λ2) incident on the second semiconductor sensor 53. The third channel EL3 represents the process emission signal in spectrum 10 visible (VIS) incident on a third semiconductor sensor 55, where the relative peak signal intensity lies between in a wavelength intermediate between the intensity peaks of the incident RL1, RL2 signals respectively on the first 52 and second 53 semiconductor sensors. The fourth channel is EL4 represents the thermal emission signal of the process in spectrum 15 infrared (IR) incident on a fourth semiconductor sensor 56. Referring to figure 11, the first semiconductor sensor 52 includes a photodiode. Similarly, the second semiconductor sensor 53 includes a photodiode. With reference to figure 10 the third sensor semiconductor 55 includes a photodiode. Similarly, the fourth sensor 20 semiconductor 56 includes a photodiode. With reference to figures 18A, 18B, 18C, 18D, the curves of PDR1, PDR2 responsiveness as a function of the wavelength of their respective first 52, second 53, third 55 and fourth 56 semiconductor sensor. In particular, in figure 18A they are represented as a function of length 25 wavelengths and superimposed: on one side the signal intensity of the first light channel reflected RL1 (dimensionless, %) incident on the first sensor 52 a semiconductor and on the other hand the responsivity (electric current generated by incident power, i.e. A / W). This first semiconductor sensor 52 will be capable of detecting signals in a predetermined band, presenting a respective lower threshold of LBT1 band in a wavelength 5 predetermined and a respective upper band threshold HBT1 in a predetermined wavelength. Similarly, Figures 18B, 18C and 18D represent as a function of the wavelength and superimposed: on the one hand the signal intensity of the respective second, third and fourth light channels RL2, EL3, EL4 (dimensionless, %) 10 accident on the respective second sensor 52, third sensor 55 and fourth sensor 56 semiconductor and on the other the PDR1, PDR2 response curves (current electrical power generated by incident power, i.e. A / W). With reference to figure 18B and according to a particular embodiment interest, the second semiconductor sensor 53 presents a first curve 15 of PDR1 responsiveness substantially equal to that of the first sensor 52 a semiconductor. Such first and second semiconductor sensors 52, 53 they could be two type 1 photodiodes with similar technical specifications. With reference to figure 18C and according to a particular embodiment interest, the third semiconductor sensor 55 presents a first curve of 20 PDR1 responsiveness substantially equal to that of the first and second semiconductor sensor 52, 53. This third semiconductor sensor 55 can be a type 1 photodiode with similar technical specifications to those of the first or second semiconductor sensor 52, 53. With reference to figure 18D and according to a particular embodiment 25 interest, the fourth semiconductor sensor 56 has a second curve of PDR2 responsiveness different from the first PDR1 responsiveness curve of the first, second and third semiconductor sensors 52, 53, 55. In such cases, the fourth semiconductor sensor 56 is a type 2 photodiode with specifications techniques different from those of the first, second or third sensor 52, 53, 55 a semiconductor. 5 As regards figure 18A, a first one is also represented threshold wavelength LPF1 of the first optical filter (low-pass filter) 54 and a third threshold wavelength LPF3 of the third optical filter (filter low-pass) 58. The concatenation of the first 54 and third 58 optical filters in combination with the intrinsic PDR1 response curve of the type photodiode 10 1 of the first sensor 52 will allow to detect the signal in a first band WB1 of wavelengths including the wavelength λ of operation of the oscillator 21 of the first laser device 2. This first WB1 band will be compressed between the first LPF1 and the third LPF3 length threshold wave. 15 Furthermore, figure 18B graphically reports a second wavelength of LPF2 threshold of the second optical filter (low-pass filter) 57. The second filter optical 57 interposed between the second sensor 53 and the light source of the second RL2 channel will appropriately filter the light with longer wavelengths at the second threshold wavelength LPF2, detecting the signal included 20 in a second WB2 band of wavelengths between the lower band threshold LBT1 of the type 1 photodiode and the second threshold wavelength LPF1. With reference to figure 18C, the reasoning set out above is reiterated, in which the The first 54 and second 57 filters have wavelengths of LPF1 and LPF2 threshold. The concatenation of the first 54 and second 57 filters in 25 combination with the intrinsic PDR1 response curve of the type photodiode 1 of the third sensor 55 will allow to detect the signal in a third band WB3 of wavelengths in the visible spectrum. With reference to figure 18D the third wavelength is also represented threshold LPF3 of the second optical filter (low-pass filter) 58. The third optical filter 5 58 interposed between the fourth sensor 56 and the light source of the fourth channel EL4 will appropriately filter light with wavelengths shorter than the third threshold wavelength LPF3, detecting the signal included in a fourth WB4 band of wavelengths compressed between the third wavelength of threshold LPF3 of the third optical filter 58 and the upper band threshold HBT2 of the 10 type 2 photodiode. With reference to figures 16, 19A-19C, 20A-20C, 21A-21C, 22A-22C, they are represent the components of the time series as a function of time light intensity signal relating to the incidence of the main beam B1 of the first laser device 2, working in the first wavelength λ1 and of the 15 auxiliary beam B2 of the second laser device 3 operating in the second wavelength λ1. In particular, with respect to the signals recorded by the respective first semiconductor sensor 52, second semiconductor sensor 53, third semiconductor sensor 55 and fourth semiconductor sensor 56, in respective first WB1, second WB2, third WB3 and fourth WB4 light spectra. 20 In particular, in fig. 19A the component 61 of the series is represented temporal of the first retro-reflection signal RL1 due to the incidence of the main beam B1 on artifact W. 19B represents component 62 of the time series of the first retro-reflection signal RL1 due to at the incidence of the auxiliary beam B2 on the artifact W, which in this case is 25 negligible. With reference to fig. 19C, the sum of the two components 61, 62 in the first wavelength λ1 constitutes the time series 63 of the first RL1 retroreflection signal recorded by the first semiconductor sensor 52. Fig. 20A represents component 71 of the second time series retro-reflection signal RL2 due to the incidence of the main beam B1 5 on the W artifact; in this case negligible. 20B represents the component 72 of the time series of the second retroreflection signal RL2 due to the incidence of the auxiliary beam B2 on the artifact W. With referring to fig. 20C, the sum of the two components 71, 72 in the second wavelength λ constitutes the time series 73 of the second signal of 10 retro-reflection RL2 recorded by the second semiconductor sensor 53. Fig. 21A represents component 81 of the third time series EL3 plasma signal due to the incidence of the main beam B1 on the artifact W, while fig. 21B represents component 82 of the series temporal delay of the third EL3 plasma signal due to the beam incidence 15 auxiliary B2. With reference to fig. 21C, the sum of the two components 81, 82 in the third visible light spectrum WB3 constitutes the time series 83 of the third plasma signal EL3 recorded by the third semiconductor sensor 55. Fig. 22A represents component 91 of the quarter time series. EL4 thermal infrared signal due to the incidence of the main beam B1 20 on artifact W, while fig. 22B represents component 92 of the series temporal variation of the fourth thermal infrared signal EL4 due to the incidence of the auxiliary bundle B2. With reference to fig. 22C, the sum of the two components 91, 92 in the fourth thermal infrared spectrum WB4 constitutes the time series 93 of the fourth thermal infrared signal EL4 recorded by 25th semiconductor sensor 56. With reference to fig. 23, the time series 63, 73 are represented detected by the respective first semiconductor sensor 52 and the respective second semiconductor sensor 53 in a single graph, where they are defined, for a predetermined welding program at least an initial time tstart 5 coinciding with the start of the pre-fusion stage ∆ start, a start time tmelt of the metal melting stage ∆melt, a start time tsolid of the stage of ∆solid cooling and a final welding cycle time tend. Each welding program will present default time values ​​of different welding stages, so that the time series 63, 73, 83, 93 of the RL1 signals, 10 RL2, EL3, EL4 detected by their respective semiconductor sensors 52, 53, 55, 56 can be cut into a plurality of segments to be analyzed separately. With reference to fig. 24, the values ​​are also represented graphically averages 66, 76 of the respective time series 63, 73 real of the respective signals 15 RL1, RL2 of retro-reflection. The calculation of the average value 66, 76 is made by summing or integrating the values ​​of the time series 63, 73 and divided by the reference interval time. The calculation of the average value 66, 76 is done by summing or integral of the time series values ​​63, 73 and divided by the time 20 of the reference range. For example, if we take the entire duration of a period as a reference interval welding cycle, the average value will be calculated considering the initial time tstart, before the first and second laser devices 2, 3 start emitting the respective main beams B1 and auxiliary beams B2 and tend, once the laser devices 25 2, 3 no longer emit their respective beams B1, B2 at the end of the stage ∆solid cooling of the solder joint. It is possible to define, for a predetermined welding program, a predetermined acceptability band with an upper threshold of 64, 74 and a lower threshold of 65, 75. The average values ​​66, 76 compressed within the acceptability bands defined by the thresholds 64-65, 74-75 will be considered normal. 5 In the example shown in the upper part of Figure 24, the average value 66 of the time series 63 detected by the first semiconductor sensor 52 is included within the acceptability band defined by the upper threshold 64 and the lower threshold 65. In this sense, according to the criteria of the programme welding the first retro-reflection signal RL1 (first wavelength λ ) 10 does not present any anomalous values. As for the example relating to the second retro-reflection signal RL2 (second wavelength λ ), the average value 76 of time series 73 is outside the acceptability band represented by the upper threshold 74 and lower 75, therefore, according to the welding program criteria, this would be an anomalous value to 15 evaluate. Even if not represented in the figure, the same reasoning can be applied to time series 83 and 93 of the respective third EL3 plasma signal and the fourth thermal infrared signal EL4. It is reiterated that, even if not reported in figure 25, it will be possible to apply the reasoning just explained 20 at predetermined time intervals defined by the welding program. Each of these ranges will have a predetermined acceptability band and the calculated average value will represent a physical indicator to be evaluated further, in order to detect possible anomalies and / or defects relating to the specific time interval (stage) of the welding process. 25 With reference to figures 25-27, the welding method also includes the calculation of welding quality indicators from the real signal 63, 73, 83, 93 detected by the respective semiconductor sensor 52, 53, 55, 56. The the same reasoning applies to the reference signals 67, 77, 87, 97. In general, a real signal 63, 73, 83, 93 or reference 67, 77, 87, 97 of the optical emission EL, RL coming from the artefact W can be defined 5 as a time-varying intensity I(t) at different instants of time t. In general, a physical indicator can be defined as the energy of a respective real signal 63, 73, 83, 93 or reference 67, 77, 87, 97 inside of a respective time period ti-tf, where each time period ti-tf is relative at a respective stage of the welding operation, the duration of this stage 10 varying according to the specific welding program (produced by (realize). In other words, a physical indicator is defined as the energy E of a respective portion of the real signal 63, 73, 83, 93 or reference signal 67, 77, 87, 97, each portion of the signal being related to a respective stage of the welding operation. 15 As stated, in a preferred embodiment, the controller 6 is configured to extract one or more physical indicators from the at least one real signal 63, 73, 83, 93 or reference 67, 77, 87, 97 of the optical emission RL, EL of the W. artifact Appropriately, using a first mode of extraction of the indicators 20 physical signals, the controller 6 is configured to extract each of the physical indicators by calculating the energy E of a respective real signal 63, 73, 83, 93 or reference 67, 77, 87, 97 within a respective time period ti- tf via the integral of the time-varying intensity I(t) from an instant initial time ti to a final time instant tf, according to the following 25 mathematical formula: 𝑝𝑝 𝑓𝑓 ( ) 𝑌𝑌 𝐸𝐸 = � 𝐼𝐼 𝑡𝑡 𝑑𝑑𝑡𝑡 𝑝𝑝 𝑖𝑖 Where: − “E” is the total energy of the real signal 63, 73, 83, 93 or reference 67, 77, 87, 97 within a ti-tf time period; 5 − “Y” is a variable that indicates the type of real signal 63, 73, 83, 93 or reference 67, 77, 87, 97 observed, which can be “BR1” for the first RL1 retroreflection signal, in particular the time series real 63 or reference 77, “BR2” for the second feedback signal RL2 reflection, in particular the real or reference time series 73 10 77, “P” for the EL3 plasma signal, in particular the time series real 83 or reference 87, and “T” for the thermal infrared signal EL4, in particular the real time series 93 or reference 97; − “positionX” is a variable that qualitatively indicates the period of ti-tf time, i.e. a specific stage of the welding operation, and 15 then the time domain in which the integral is calculated, in particular positionX can refer to an initial stage ∆start prior to metal melting point, ∆melt relating to the range in which the metal melts while the laser sources are operating3 and ∆solid is the range in which the solder joint on the artifact W is cooled; 20 − “I(t)” is the time-varying intensity of the real signal 63, 73, 83, 93 or of reference 67, 77, 87, 97 of the optical emission EL, RL of the artifact W; − “ti” is the instant of time in which the integral begins, or rather in which the time domain in which the integral is calculated; and 25 − “tf” is the instant of time at which the integral ends, or rather at which the time domain in which the integral is calculated. 𝑝𝑝 𝑝𝑝𝑝𝑝 𝑝𝑝 𝑝𝑝𝑝𝑝 𝑝𝑝𝑝𝑝𝑝𝑝 Figures 25-27 graphically show the extraction of physical indicators BR1start, BR1melt, BR1solid, BR2start, BR2melt, BR2solid, Pstart, Pmelt, Psolid, THstart, THmelt, THsolid of a signal, real 63, 73, 83, 93 or reference 67, 77, 87, 97, of the issue 5 As reported above, in a preferred embodiment, the controller 6 is configured to extract one or more optical physical indicators, using this first extraction mode. However, considering that the real signal 63, 73, 83, 93 or reference 67, 77, 87, 97 of the optical emission EL, RL from the W artifact 10 are preferably of digital type, the integral of the variable intensity in the time I(t) from an initial time instant ti to a final time instant tf, to calculate the energy E of a respective real signal 63, 73, 83, 93 or of reference 67, 77, 87, 97 within a respective time period ti- tf, can be approximated. 15 Therefore, advantageously, using a second mode of extraction of the physical indicators of a signal, the controller 6 is configured to extract each of the physical indicators by calculating the energy And of a respective real signal 63, 73, 83, 93 or reference 67, 77, 87, 97 within a respective time period ti-tf via summation 20 of the instantaneous intensity Ii from an initial time instant ti to an instant final time tf, multiplied by the sampling time Δt (i.e. related to the sampling frequency) signal acquisition), according to the following mathematical formula: 𝑝𝑝 𝑓𝑓 𝐸𝐸 𝑌𝑌 = ∆ 𝑡𝑡 � 𝐼𝐼 𝑝𝑝 𝑝𝑝 = 𝑝𝑝 𝑖𝑖 𝑝𝑝 𝑝𝑝𝑝𝑝 𝑝𝑝 𝑝𝑝𝑝𝑝 𝑝𝑝𝑝𝑝 𝑝𝑝 Where: − “E” is the total energy of the real signal 60-72 or reference 50- 58 within a tstart-tend time period; − “Y” is a variable that indicates the type of real signal 63, 73, 83, 5 93 or reference 67, 77, 87, 97 observed, and which can take on value “BR1” for the first retro-reflection signal RL1, in particular the series real time 63 or reference 77, “BR2” for the second signal retro-reflection RL2, in particular the real time series 73 or reference 77, “P” for the EL3 plasma signal, in particular the series 10 real time 83 or reference 87, and “T” for infrared signal EL4 thermal, in particular the real or reference time series 93 97; − “positionX” is a variable that qualitatively indicates the period of time tstart-tend, i.e. a specific stage ∆start, ∆melt, ∆solid 15 of the welding operation, and hence the time domain in which the sum is calculated; Referring to figures 26-29, together with time series 63, 73, 83, 93 real of the respective back-reflection signals RL1, RL2, plasma EL3 and thermal EL4 represents the respective time series 67, 77, 87, 97 of 20 reference. These time series 67, 77, 87, 97 correspond to the set of expected values ​​for a predetermined welding program. The controller is further configured to calculate, starting from such time series 67, 77, 87, 97 reference of the acceptability bands. In particular, starting from the time series 67, 77, 87, 97 of reference it will be possible to calculate, for 25 each value of the time series 67, 77, 87, 97 of reference a value of lower threshold and an upper threshold value. With particular reference to figures 28 and 29, for the retro-reflection signals RL1, RL2, in particular for the respective reference time series 67, 77, the respective upper threshold series 68, 78 are advantageously calculated and lower 69, 79. Even if not specifically reported graphically in the 5 figures the same reasoning is valid for the reference time series of plasma 87 and thermal 97. The controller, in fact, is configured to detect values punctual or a series of outliers of the real time series 63, 73 that fall outside the threshold series. Referring to figure 28, the values ​​of series 63 relating to the first signal 10 of RL1 retro-reflection are compressed within their respective threshold series upper 68 and lower 69. However, the values ​​of the 73 series relating to the second RL2 retro-reflection signal comes out of their respective threshold series upper 78 and lower 79. In such cases, the controller is configured to report an anomalous situation. 15 With particular reference to figure 29, the values ​​of series 73 relating to the second retro-reflection signal RL2 are compressed within the respective series of upper threshold 78 and lower threshold 79 of the respective reference signal RL2rif. However, the values ​​of the 63 series relating to the first feedback signal reflection RL1 come out of their respective series of upper threshold 78 and 20 lower 79. In particular, a plurality of values ​​of the real time series 63 it goes beyond the upper threshold 68 in a particular time interval 631 during the metal melting stage. Controller 6 will be configured to identify the type of anomaly based on the threshold exceeded and the stage in which it happens. 25 With reference to figure 30, a summary of the table is presented correlation between the different possible defects of a soldering station 1 with two laser devices 2, 3 and their respective signals RL1, RL2, EL3, EL4 monitored. Ad example, monitoring the first RL1 and second RL2 feedback signal reflection will allow to identify situations of welding instability (spatter or splash phenomenon due to implosion of the welding bubble) when 5 the second signal RL2 reports higher values ​​than the reference signal RL2rif. Furthermore, the welding method provides for a comparison between the feedback signals reflection RL1, RL2 recorded and the expected values. The calculation of the expected values ​​can be achieved by calculating the respective threshold bands for the signals retro-reflection RL1, RL2. The threshold band relating to the first retro-reflection signal 10 reflection RL1 will consider the operating parameters of the first device laser 2, with which to calculate the emission of the main beam B1 in the first wavelength λ1 and the reflectance of the material on which the beam is incident main B1 in that first wavelength λ1. The threshold band relating to the second retro-reflection signal RL2 will consider the parameters of 15 operation of the second laser device 3, with which to calculate the emission of the auxiliary beam B2 in the second wavelength λ2 and the reflectance of the material on which the auxiliary beam B2 is incident at this second wavelength λ2. For example, the fluctuation of the first RL1 and / or second RL2 feedback signals reflection outside the respective threshold bands will indicate the presence of 20 gaps. It is clear that the station and method for laser welding described so far changes and / or additions of parts or phases can be made, without this goes beyond the scope of this invention as defined by In the following claims, the references in parentheses are for convenience only. to facilitate reading and should not be considered as limiting factors for as regards the scope of protection underlying the specific claims.

Claims

claims 1. Laser welding station comprising: - a first laser device (2) comprising a respective oscillator (21) configured to emit a main beam (B1) in a first wavelength (ài), wherein the first laser device (2) is preferably configured to emit said main beam (B1) in the near infrared (NIR) spectrum; - a second laser device (3) comprising a respective oscillator (31) configured to emit an auxiliary beam (B2) in a second wavelength (À2) different from the first wavelength (ài), wherein the second laser device (3) is preferably configured to emit said auxiliary beam (B2) in the visible light (VIS) spectrum, in particular in the blue light spectrum;- an optical head (4) configured to emit a laser beam (LB) so as to create a well of molten metal to join together at least two elements (W1, W2) of an artefact (W), said laser beam (LB) being obtained by combining together said main beam (B1) and said auxiliary beam (B2); - first (23) and second (33) optical organs configured to transmit the respective first (B1) and second (B2) beams to said optical head (4); and - a monitoring device (5) configured to detect an emission of reflected light (RL) from the artefact (W) being processed; characterized in that said monitoring device (5) comprises: - A first semiconductor sensor (52), configured to detect reflected light (RL) in a first wavelength spectrum (WB1) substantially equal to the first wavelength (λι) of the main beam (B1), thereby monitoring the absorbance of the main beam (B1) by the artefact (W);and - A second semiconductor sensor (53), configured to detect reflected light (RL) in a second wavelength spectrum (WB2) substantially equal to the second wavelength (λ2) of the auxiliary beam (B2), thereby monitoring the absorbance of the auxiliary beam (B2) by the artifact (W).; 2. Laser welding station according to claim 1, characterized in that said monitoring device (5) further comprises a monitoring mirror (51) configured to reflect the light of the main beam (B1) and the auxiliary beam (B2) of the laser beam (LB) and the reflected light (RL) from said component (W) towards at least said first (52) and said second (53) semiconductor sensors.

3. Laser welding station according to claim 2, characterized in that said monitoring device (5) further comprises a first optical filter (54) interposed between said monitoring mirror (51) and said first (52) and second (53) semiconductor sensor, said first optical filter (54) being configured to reflect at least a second back-reflection signal (RL2) of the reflected light (RL) towards said second semiconductor sensor (53) and to transmit at least a first back-reflection signal (RL1) of the reflected light (RL) towards said first semiconductor sensor (52).

4. Laser welding station according to any of the preceding claims, characterised in that said monitoring device (5) further comprises a third semiconductor sensor (55), configured to detect the light emitted by the welding process (EL) in a third visible light spectrum (WB3), so as to monitor the plasma of the molten metal well generated by the incidence of the main beam (B1) and the auxiliary beam (B2) on the artefact (W).

5. Laser welding station according to claim 3 or 4, characterized in that said monitoring device (5) further comprises a second optical filter (57) interposed between said first optical filter (54) and said second semiconductor sensor (53) and configured to reflect said second retro-reflection signal (RL2) of the reflected light emission (RL) with a wavelength spectrum substantially coinciding with the second wavelength (À2) emitted by said second laser device (3).

6. A laser welding station according to claim 5 when dependent on claim 4, characterized in that said second optical filter (57) is further configured to transmit a third plasma signal (EL3) of the emitted light (EL) in the third visible light spectrum (WB3) with wavelengths longer than those of the second spectrum (WB2), said second optical filter (57) being arranged so as to orient said third plasma signal (EL3) towards said third semiconductor sensor (55).

7. Laser welding station according to any of the preceding claims, characterized in that said monitoring device (5) further comprises a fourth semiconductor sensor (56) configured to detect the light emitted (EL) by the welding process in a fourth thermal infrared spectrum (WB4) so ​​as to monitor the temperature of the artefact (W), the temperature of the respective molten metal well formed on it and the temperature during subsequent cooling once the main (B1) and / or auxiliary (B2) beams are turned off.

8. Laser welding station according to claim 7, characterized in that said monitoring device (5) further comprises a third optical filter (58) interposed between said first optical filter (54) and said first semiconductor sensor (52) and configured to reflect a first retro-reflection signal (RL1) of the reflected light emission (RL) with a wavelength spectrum substantially coinciding with the first wavelength (ài) emitted by said first laser device (2).

9. Laser welding station according to claim 8 when dependent on claim 7, characterized in that said third optical filter (58) is further configured to transmit a fourth thermal infrared signal (EL4) emitted by the welding process in the fourth thermal infrared spectrum (WB4), with wavelengths longer than those of the first light spectrum (WB1), said third optical filter (58) being arranged so as to orient said fourth thermal infrared signal (EL4) towards said fourth semiconductor sensor (56).

10. Laser welding method for welding together at least two elements of an artefact, wherein said method comprises the steps of: a. Arranging at least two elements (W1, W2) of an artefact (W) close together in a welding area and facing an optical head (4) of a laser welding system (10); b. Scanning at least one of the at least two elements (W1, W2) of said artefact (W) with a welding laser beam (LB) to create a well of molten metal on itself, wherein the welding laser beam (LB) is formed by at least a main beam (B1), emitted by a first laser device (2) at a first wavelength (ài), wherein the first wavelength (ài) is preferably in the near infrared (NIR), and an auxiliary beam (B2), emitted by a second laser device (3) at a second wavelength (À2), wherein the second wavelength (À2) is preferably in the visible spectrum (VIS), in particular in the blue-light spectrum; c.Cooling the molten metal well by turning off the welding laser beam (LB) to obtain a solidified weld joint; d. Monitoring a reflected light emission (RL) during said scanning and cooling phases to evaluate the quality of said weld joint, characterized in that the monitoring phase is composed of the following sub-phases: d.1 ) monitoring a first back-reflection signal (RL1) of the reflected light emission (RL), wherein said first back-reflection signal (RL1) is formed by an electromagnetic emission in a first spectrum (WB1) of wavelengths substantially coincident with the first wavelength (ài) emitted by said first laser device (2); and d.2 ) monitoring a second back-reflection signal (RL2) of the reflected light emission (RL), wherein said second back-reflection signal (RL2) is formed by an electromagnetic emission in a second spectrum (WB2) of wavelengths substantially coincident with the second wavelength (λ2) emitted by said second laser device (3).

11. Method according to claim 10, characterized in that the sub-phase of monitoring (d.1) a first retro-reflection signal (RL1) of the reflected light emission (RL) consists in calculating a first tolerance band of the signal recorded by the first semiconductor sensor (52) based on the operating parameters of the first laser device (2) of the relevant scanning phase and on the reflectance of the material on which the main beam (B1) in the first wavelength (ài) is incident, and in that the sub-phase of monitoring (d.2) a second retro-reflection signal (RL2) of the reflected light emission (RL) consists in calculating a second tolerance band of the signal recorded by the second semiconductor sensor (53) based on the operating parameters of the second laser device (3) of the relevant welding program and on the reflectance of the material on which the auxiliary beam (B2) in the second wavelength (À2) is incident.

12. A method according to claim 10 or 11, characterized in that said monitoring step further comprises the sub-step of d.3) monitoring a third plasma signal (EL3) of the emitted light (EL) due to the scanning step, wherein said third plasma signal (EL3) is formed by an electromagnetic emission in a third visible light (VIS) spectrum (WB3). and in that said monitoring step further comprises the sub-step of d.4) monitoring a fourth thermal infrared signal (EL4) of the emitted light (EL) of the welding process, wherein said thermal infrared signal (RL4) is formed by an electromagnetic emission in a fourth infrared spectrum (WB4) and different from the first wavelength (ài).

13. Method according to one of claims 10-12, characterized in that each of said monitoring sub-phases (d.1, d.2, d.3, d.4) comprises the steps of: - detecting and acquiring said signals (RL1, RL2, EL3, EL4) coming from said artifact (W) for the duration of said scanning and cooling phases (b, c) of each welding operation between said elements (W1, W2) of said artifacts (W); - obtaining at least the real signals (RL1, RL2, EL3, EL4) relating to the respective light spectra (WB1, WB2, WB3, WB4) from said reflected and emitted signal (RL, EL), said real signals (RL1, RL2, EL3, EL4) being a time trend consisting of one of a continuous time series and a plurality of consecutive discrete values ​​of a discrete time series;- comparing characteristics of the at least one time trend (63, 73, 83, 93) of the signals (RL1, RL2, EL3, EL4) with respective characteristics of at least one time trend (67, 77, 87, 97) of a respective reference signal, identifying defects in the welding process of the artifact (W) based on the relationship between said compared characteristics; and - classifying said real signals (RL1, RL2, EL3, EL4) based on the comparison of said characteristics of said real signals (RL1, RL2, EL3, EL4) with said respective characteristics of said reference signals, identifying any defects and / or estimating mechanical properties of the welding of said artifact (W) based on the unique relationship between said compared characteristics.; 14. A method according to claim 13, characterised in that prior to said scanning (b.) and cooling (c.) steps, said method comprises the step of: - selecting a predefined welding program, where said welding program comprises operating instructions and parameters with which a controller (6) executes said scanning and cooling steps, where said welding program further comprises at least one reference signal (RLIrif, RL2rif, EL3rif, EL4rif) of said reflected signal (RL) relating to said welding program, and where said welding program further comprises at least one monitoring time window (ti-tf) in said scanning and cooling steps, such time window (ti-tf) corresponding to a stage (tstart, Δmelt, Δsolid) of said welding operation with initial time (tstart, tmelt, tsolid) and final time (tmelt, tsolid, tend);in that said monitoring step further comprises, before said comparing step, the steps of: - extracting values ​​relating to said stage (^start, ΔπβΚ, Δsolid) defined by said at least one time window (ti-tf) from said at least one real signal (RL1, RL2, EL3, EL4) and from said at least one respective reference signal (RL1rif, RL2rif, EL3rif, EL4rif) of each welding operation; and - calculate at least one physical indicator (BR1start, BR2start, Pstart, Tstart, BR1melt, BR2melt, Pmelt, Tmelt, BR1 solid, BR2solid, Psolid, Tsolid) relating to said stage (^start, ΔπβΚ, Δsolid) of said welding operation, starting from said values ​​enclosed in said at least one time window (ti-tf) and extracted from said at least one real signal (RL1, RL2, EL3, EL4) and from said at least one respective reference signal (RL1 ref, RL2rif, EL3rif, EL4rif);and in that said step of comparing characteristics of the at least one time trend further comprises the step of: - comparing said at least one physical indicator (BR1 start, BR2start, Pstart, Tstart, BR1melt, BR2melt, Pmelt, Tmelt, BR1 solid, BR2solid, Psolid, Tsolid) of said stage pstart, ΔπβK, Δsolid) of said at least one real signal (RL1, RL2, EL3, EL4) with said at least one respective physical indicator of said same stage ^start, Δmelt, Δsolid) of said at least one reference signal (RL1rif, RL2rif, EL3rif EL4rif).;