Roll forming machine with sensor roll forming frame and method for roll forming or bending in the roll forming machine

Abstract

Roll forming plant for roll forming a sheet metal strip (9) with a plurality of roll forming stands arranged in series, which form and / or guide the sheet metal strip (9) in succession, comprising: a) at least one sensor roll forming stand (5) arranged in the plurality of roll forming stands arranged one behind the other and which forms, bends and / or guides the sheet metal strip (9); b) a first roller and a second roller which are rotatably connected to the respective sensor roll forming frame (5) about a respective axis of rotation and are held in position by it, the first and second roller being arranged so that they clamp the sheet metal strip (9) between them; c) at least one sensor (1, 2) arranged in the respective sensor roll forming stand (5) in such a way as to measure a force, vibration or torque exerted by the respective sensor roll forming stand (5) on the sheet metal and to generate a corresponding sensor signal (S1); characterized by d) a microprocessor unit (6) with a detector algorithm configured to compare the sensor signal (S1) with a corresponding predetermined setpoint or setpoint curve within a corresponding predetermined permissible tolerance range, wherein, if the sensor signal (S1) is detected as being within the permissible tolerance range, a first state and, otherwise, at least a second state of the respective sensor roll forming frame (5) is detected; and e) a display unit (8) that displays the detected first and / or second state for the respective sensor roll forming frame (5), f) wherein the detector algorithm is configured to sample, store and statistically evaluate the sensor signal (S1) over a predetermined period in response to an operator signal in order to determine and store the setpoint or setpoint curve and / or the permissible tolerance range.

Description

[0001] Roll forming machine with sensor roll forming stand and process for roll forming or bending in the roll forming machine.

[0002] The present invention relates to a roll forming system according to the preamble of claim 1 and a method according to the preamble of claim 12.

[0003] Conventional roll forming systems deform a sheet metal strip into a desired sheet profile in successive steps using a multitude of roll forming stands arranged one behind the other. Each roll forming stand makes a small adjustment to the sheet metal strip, or guides or drives it at a specific location. The roll forming stands must be aligned as precisely as possible to each other and perform a predetermined deformation as precisely as possible.

[0004] DE 10 2007 062 104 A1 discloses a conventional roll-forming system with a plurality of successively arranged roll-forming stands that process the sheet metal strip step by step. At the end of the roll-forming system, the resulting sheet metal profile is typically measured at periodic intervals to ensure compliance with predetermined dimensions and permissible tolerances, or to reject the sheet metal profile if it is not. This almost inevitably results in scrap, and the roll-forming stands must be readjusted if predetermined tolerances are exceeded. Early detection of a deviation is not possible with the disclosed roll-forming system.

[0005] DE 10 2008 036 275 B4 discloses an optical sensor arranged in or after the roll forming machine for measuring the sheet metal profile in order to ensure compliance with the predetermined dimensions or to cause rejection.

[0006] DE 10 2015 223 516 A1 discloses a rolling mill with a rolling stand and a method for actively damping vibrations in the rolling stand, wherein a backup roll for supporting a work roll acting on the sheet metal strip is designed to be adjustable in the rolling stand.

[0007] US 5 761 945 A discloses another rolling mill with a rolling stand in which a rolling contact pressure force in the rolling stand is measured and controlled by sensors.

[0008] DE 36 41 748 A1 discloses a rolling mill and a method for measuring and controlling a degree of deformation.

[0009] DE 10 2006 027 066 A1 discloses a rolling mill and a method for guiding the sheet metal strip.

[0010] When setting up a roll forming line, there is inevitably a considerable amount of waste at the beginning until all the rolls or roll forming stands are correctly adjusted to each other and to the sheet metal strip.

[0011] Roll forming lines also often have roll forming stands distributed over a length of 3 meters or more, so that an adjustment of an initial roll forming stand, for example, forming a first edge, is only detected at the end, resulting in scrap along the entire length of the roll forming line. Since scrap is associated with unnecessary costs, it is of course important to keep this to a minimum.

[0012] In the event of a warpage or, for example, a bearing failure of a roll or cylinder in a roll forming stand, any resulting changes in the sheet metal profile at the end of the roll forming line are either detected immediately or only during a periodic inspection. The roll forming line operator must then investigate the actual cause of the change or defect. A part is then either replaced or adjusted. This results in a delay, and the line must be operated by a highly trained specialist—all of which are cost factors that should be kept as low as possible.

[0013] The object of the invention, in order to eliminate the disadvantages of the prior art, is to provide a device and a corresponding method which detects and indicates deviations from a target profiling process as early as possible

[0014] The above object is achieved by a device according to the features of independent claim 1 or a method according to the features of independent claim 12. Further advantageous embodiments of the invention are specified in the dependent claims.

[0015] According to the invention, a roll forming plant for roll forming a sheet metal strip with a plurality of roll forming stands arranged one behind the other, which form and / or guide the sheet metal strip in sequence, and a corresponding method are provided, wherein the roll forming plant comprises the following: - at least one sensor roll forming stand arranged in the plurality of roll forming stands arranged one behind the other and which forms, bends and / or guides the sheet metal strip; - a first roller and a second roller connected to the respective sensor roll forming stand for rotation about a respective rotation axis and held in position thereby, the first and second rollers being arranged to clamp the sheet therebetween; - at least one sensor arranged in the respective sensor roll forming stand to measure a force, a vibration or a torque exerted by the respective sensor roll forming stand on the sheet and generate a corresponding sensor signal; - a microprocessor unit with a detector algorithm designed to compare the sensor signal with a corresponding predetermined target value or a target value curve at a corresponding predetermined permissible tolerance range, wherein if the sensor signal is detected to be within the permissible tolerance range, a first state and otherwise at least a second state of the respective sensor roll forming stand is detected; and - a display unit that displays the detected first and / or second state for the respective sensor roll forming stand.

[0016] The roll forming system according to the invention with the sensor roll forming stand allows, after determining the setpoint or the setpoint curve and the tolerance range, for example, in a test run that is running properly, a deviation of the sensor signal from the tolerance range to be immediately detected during operation, thereby identifying a process deviation. Upon detection of the second state or a further state, conclusions can be drawn, for example, based on typical patterns in the sensor signal, regarding specific damage, specific wear, a change in sheet thickness, a profile curve that deviates from a predetermined profile curve, and other influences. Another particularly advantageous feature is that preferential pattern recognition in the sensor signal or in the sensor signal curve can be used to establish a preventative diagnosis for the system.For example, bearing damage or roll damage can be detected early, allowing replacement parts or a specialist to be ordered without having to completely stop the roll forming line. A change in sheet thickness or a process change in one of the preceding roll forming stands can also be detected. The sensor signal can also be used to adjust preceding roll forming stands, for example, to press the sheet metal strip more flatly or similarly.

[0017] When detecting the sensor signal, the sensor signal can also be displayed as a periodic signal curve, such as in Fig. 3, whereby the sensor signal is compared with a setpoint curve or a curved tolerance range.

[0018] Through early detection or preventive diagnosis, scrap can be significantly reduced and service intervals can be adjusted to the actual situation. The corresponding cost savings are evident to the specialist.

[0019] Various methods for determining the setpoint, the setpoint curve, and the tolerance range are listed in the detailed description. It is particularly noteworthy that learning pattern recognition allows problematic conditions on the sensor roll forming stand to be detected and assigned much more effectively, allowing early diagnosis for adjustment or replacement of a spare part. The pattern recognition or the detector algorithm, which can comprise a multitude of detection algorithms, can be "taught" via an operator signal. Advantageously, the detector algorithm detects the condition and, if appropriate, any wear or adjustment of the sensor roll forming stand, and displays this to the user or operator via the display unit. It is also useful for the detector algorithm to determine a condition trend in order to detect wear, for example, of a bearing, at an early stage.The operator can also be given instructions on how to restore the correct state. Such learning pattern recognition algorithms are known in the state of the art from other applications in signal analysis or signal processing. For example, neural networks, Kalman filters, and other predictive filter algorithms can be used. Preferably, the microprocessor unit can also perform statistical evaluations based on the sensor signal and operator signals, which allow for better detection of subsequent states. Preferably, the state of the respective sensor roll forming stand is also logged by saving the sensor signal and the operator signals.Likewise, the status of each roll forming stand can be forwarded to a higher-level control system, for example to display and document the status of the respective roll forming stands and / or other parameters from a central location. Data on the status and / or sensor signals can also be transmitted, collected, and preferably evaluated via a network, for example to optimize the detection algorithm from a computer connected to the network and to transmit an optimized detection algorithm or corresponding parameters back to the microprocessor unit. Such optimization of the detection algorithm can, for example, initially improve the determination of the setpoint and / or setpoint curve, the tolerance range, and the detection parameters, making it easier for the operator.

[0020] Preferably, the microcontroller uses the detector algorithm to evaluate the sensor signal or the corresponding signal components of the respective sensor roll forming stand to which the detector algorithm is assigned. However, it is also conceivable for the detector algorithm to evaluate signal components from one or more sensors from adjacent sensor roll forming stands or from one or more other sensors arranged in the roll forming system, and preferably to determine a state of alignment of the roll forming stands relative to one another.

[0021] Preferred embodiments according to the present invention are illustrated in the following drawings and in a detailed description, but they are not intended to limit the present invention exclusively thereto.

[0022] It shows Fig. 1 schematically shows a perspective view of a section of a roll forming system for processing a sheet metal strip, wherein two adjacent roll forming stands are equipped with sensors, each generating a sensor signal component to a sensor signal, wherein the roll forming stands each comprise a first shaft for a first roll or cylinder at the top and a second shaft for a second roll or cylinder at the bottom in order to form the sheet metal strip therebetween, wherein the sheet metal strip and the cylinders or cylinders are omitted from the illustration; Fig. 2 a schematic block diagram with a sensor, a central microprocessor unit with microprocessor and memory, an operating signal generator and a display unit; Fig. 3 the sensor signal as a curve over time, showing a periodic typical pattern, and a tolerance range is shown in dashed lines; and Fig. 4 a result calculation of a finite element analysis, which represents the forces occurring as calculated vectors at a point on the sheet metal strip where a forming process is carried out by one of the roll forming stands. Detailed description of implementation examples

[0023] In Fig. 1 shows two adjacent preferred roll forming stands as a section of a preferred roll forming system for roll forming a sheet metal strip. In principle, the preferred roll forming system for roll forming the sheet metal strip typically comprises a plurality of roll forming stands arranged one behind the other, which form and / or guide the sheet metal strip in sequence. In the illustration in Fig. 1, it should be noted that the two roll forming stands are each shown with only an upper first shaft 3 and a lower second shaft 4, each of which normally carries a designated roller or roll that forms, bends, and / or guides the sheet metal strip therebetween. Accordingly, a first roller, which can also be a first roll, is preferably mounted on the first shaft 3, and a second roller, which can also be a second roll, is mounted on the second shaft 4. When the sheet metal strip is formed, bent, and / or guided between the first and second rolls, corresponding forces occur on the first and second rolls.

[0024] According to the invention, the roll forming system presented here comprises the following: a) at least one sensor roll-forming stand 5, which is arranged in the plurality of roll-forming stands arranged one behind the other and which forms, bends, and / or guides the sheet metal strip. The respective sensor roll-forming stand 5 can be arranged at the beginning, at the end, or between the plurality of roll-forming stands; b) the first roller and second roller, which are connected to the respective sensor roll-forming stand 5 so as to be rotatable about a respective rotation axis and are held in position thereby, wherein the first and second rollers are arranged such that they clamp the sheet metal strip therebetween. The respective sensor roll-forming stand 5 is preferably arranged at substantially the same height as a directly adjacent roll-forming stand in order to avoid creating undesirable tensions between the sheet metal strip and the adjacent roll-forming stand; c) at least one sensor 1, 2 arranged in the respective sensor roll forming stand 5 so as to measure a force, a vibration or a torque exerted by the respective sensor roll forming stand 5 on the sheet and generate a corresponding sensor signal; d) a microprocessor unit 6 with a detector algorithm designed to compare the sensor signal with a corresponding predetermined target value or a target value curve at a corresponding predetermined permissible tolerance range, wherein if the sensor signal is detected as being within the permissible tolerance range, a first state and otherwise at least a second state of the respective sensor roll forming stand 5 is detected; and a display unit 8 which displays the detected first and / or second state for the respective sensor roll forming stand 5.

[0025] Preferably, the at least one sensor or one of the sensors is a force sensor, such as a load cell, a strain gauge, or a current sensor, to determine a motor drive torque. Preferably, the at least one sensor or one of the sensors is a vibration sensor, a structure-borne sound microphone, or a microphone, for example, to detect vibrations on the sensor roll-forming stand 5 or in the roll-forming system and to be able to evaluate them using the detector algorithm. A temperature sensor is also conceivable as one of the sensors to detect heating, for example, due to friction.

[0026] In Fig. 1, the two roll forming stands are each shown as a sensor roll forming stand 5. In the respective sensor roll forming stand 5, two first sensors 1, 1' are preferably arranged in the sensor roll forming stand 5 such that they together measure a first force component in the Y direction, which is introduced from the sheet metal strip perpendicularly onto the first roll and the first shaft 3. Preferably, the two first sensors 1, 1' are arranged at respective ends of the first shaft 3 in such a way that together they can measure a total force on the first shaft. Fig. 1, the first sensors 1, 1' are shown as load cells, but they can also be other force sensors, such as strain gauges, which are attached to the sensor roll-forming frame 5 in such a way as to measure the first force component in the Y direction. The sensor signal would comprise at least two signal components from the first two sensors 1, 1'. The detector algorithm can either evaluate the two signal components directly and / or a resulting sensor signal, which is, for example, a sum signal of the two signal components.

[0027] At least one sensor can also be arranged in the sensor roll-forming stand 5, which measures a second force component in the X direction, which is introduced tangentially from the sheet metal strip onto the first roll and transversely to a longitudinal direction of the sheet metal strip. Furthermore, at least one sensor can also be arranged in the sensor roll-forming stand 5, which measures a third force component in the Z direction, which is introduced tangentially in the longitudinal direction from the sheet metal strip onto the first roll, for example, by a drive or a braking torque of the first shaft 3 and / or the second shaft 4.

[0028] Two second sensors 2, 2' can also be arranged in the preferred sensor roll-forming stand 5, which together determine a fourth force component in the Y direction, which is introduced perpendicularly from the sheet metal strip onto the second roll and the second shaft. The sensor signal would also include the signal components of the two second sensors 2, 2'. Likewise, the detector algorithm can evaluate both a fourth resulting sensor signal from the two signal components of the second sensors 2, 2' and each of the signal components individually.

[0029] Preferably, one or more sensors can also be arranged such that they measure a further force or corresponding force components with which the sensor roll-forming stand 5 is held in position on a machine bed in the roll-forming system, thereby generating at least part of the sensor signal. The further force can serve as a measure of the distortion of the respective sensor roll-forming stand 5 relative to an adjacent roll-forming stand, for example, if the total force on the first shaft 3 were to deviate too much from the total force on the second shaft 4, with the sheet metal in between.

[0030] Likewise, the at least one sensor in the sensor roll-forming stand 5 can also be arranged to measure a fifth force component in the X direction, which is introduced by the sheet metal strip tangentially onto the second roll and transversely to the longitudinal direction of the sheet metal strip. And the at least one sensor in the sensor roll-forming stand 5 can also be arranged to measure a sixth force component in the Z direction, which is introduced by the sheet metal strip tangentially in the longitudinal direction onto the second roll.

[0031] Further sensors and arrangements for measuring a force and / or a torque or multiples thereof are also conceivable.

[0032] The force or forces are, for example, a measure of a roll gap in relation to the thickness of the sheet metal strip, of a degree of deformation, of a material quality or hardness of the sheet metal strip, and of discontinuities in the sheet metal strip.

[0033] For clarity, it should be noted that the microprocessor unit 6, as shown in Fig. 2, comprises at least one microprocessor 6a and a memory 6b. The sensor signal can be supplied to the microprocessor unit 6 as a digital signal or as an analog signal via an analog-to-digital converter.

[0034] For the sake of clarity, it should be noted that the sensor signal can be understood as a single signal or as a plurality of individual signals or signal components that the sensor signal comprises, as is the case, for example, with a vectorial or multidimensional sensor signal. In the case of a multidimensional sensor signal, the setpoint, the setpoint curve, and / or the tolerance range are preferably also multidimensional. However, one or more resulting sensor signals can also be formed from the multiple signal components of the multidimensional sensor signal by the microprocessor unit or the detector algorithm, so that the number of setpoints, setpoint curves, and / or tolerance ranges or their components does not necessarily have to correspond to the number of sensor signal components.For example, several processed signal components can be extracted from a single sensor signal, each of which is then compared with a corresponding tolerance range or pattern recognition, as is the case, for example, with spectral analysis and subsequent pattern recognition. For clarity, it should also be noted that the sensor signal, as shown in . Fig. 3, can also be understood as a periodic sensor signal with a typical sensor signal profile. The detector algorithm then preferentially compares the periodic sensor signal with a non-constant tolerance range. Fig. Figure 3 shows, by way of example, a non-constant tolerance range between an upper threshold curve L1 and a lower threshold curve L2, between which the sensor signal may move in the first state. It is clear to those skilled in the art that in order to properly compare the sensor signal with the tolerance range, the detector algorithm must first perform period detection of the sensor signal in order to then correctly assign the tolerance range. For a simpler detection by the detector algorithm without period detection, the tolerance range would have to be selected wider.

[0035] The first state of the sensor roll forming stand 5 is preferably a preferably proper state in which the sensor roll forming stand 5 is preferably running properly in terms of process. The second state detected by the detector algorithm represents a deviation from the preferably proper state of the sensor roll forming stand 5 or of the process in the roll forming system and can be caused, for example, by a sheet metal strip that is too thick or too thin, or by initial bearing damage, or by another process deviation from the target process. The detector algorithm is preferably also designed to detect a third state which, for example, indicates an even greater or smaller deviation from the first state compared to the second state. The third state can be caused, for example, by an even greater or smaller sheet metal strip thickness or by initial or serious bearing damage.By detecting the second or a further condition that deviates from a preferred proper condition on the sensor roll forming stand, the process on the sensor roll forming stand can be stopped, an alarm signal can be issued, or an automatic actuator adjustment of the sensor roll forming stand can be carried out.

[0036] The alarm signal can generally be a visual, acoustic, or functional alarm signal. A functional alarm signal, for example, is an automatic stop or shutdown, or an actuator-based setting of the roll forming system to a safe and / or non-self-damaging state.

[0037] The display unit 8 can provide the operator with a simple pass or fail or green / red signal or similar. The display unit 8 can also provide the operator with an indication of the likely cause of the current condition, for example, bearing damage on the first shaft 3, bearing damage on the second shaft 4, sheet metal strip too thick, insufficient feed from the preceding roll forming stands, or similar. The display unit 8 indicates the condition of the respective sensor roll forming stand 5, but it can also separately display the condition of another sensor roll forming stand 5. The display unit preferably comprises a traffic light, a display, a monitor, a smartphone display, a tablet, an alarm light, a visual alarm indicator, an acoustic alarm, or a combination thereof.

[0038] In order to determine the respective tolerance range for the sensor signal or its respective signal component, the detector algorithm is preferably designed to determine and store the sensor signal as the predetermined setpoint value or setpoint curve in response to an operator signal. The operator signal can also be activated multiple times by the operator, whereby the detector algorithm then calculates, for example, an average value, such as an arithmetic or median value, to obtain the setpoint value or setpoint curve. It is advisable to generate the setpoint curve instead of a single setpoint, for example, when the sensor signal has a typically periodic characteristic, as in Fig. 3 shown.

[0039] The detector algorithm is designed to sample, store, and statistically evaluate the sensor signal over a predetermined period of time in response to an operator signal, whereby the predetermined setpoint value or the setpoint curve and / or the permissible tolerance range determined in this process is determined and stored. The detector algorithm preferably determines the predetermined setpoint value over the predetermined period of time as an arithmetic mean, a median value, or another mean value.

[0040] Preferably, the detector algorithm determines the predetermined permissible tolerance range over the predetermined period of time by detecting minimum and maximum values ​​with an additional tolerance offset or factor offset. The tolerance offset extends the tolerance range and the offset value(s). A factor offset is understood to be an offset resulting from multiplying a factor by a previously determined tolerance value.

[0041] Preferably, the target value is statistically determined as the mean value and the tolerance range, for example as a calculated variance multiplied by a predetermined factor, from a series of stored sensor data during a test run.

[0042] Preferably, the detector algorithm can determine the predetermined setpoint curve and the permissible tolerance range over the predetermined period of time by detecting a periodicity in the sensor signal using a known period detection algorithm and thus decomposing the sensor signal into a series of periodic sensor signal curves. The period detection algorithm determines an averaged curve from the series of periodic sensor signal curves as the predetermined setpoint curve, with the permissible tolerance range being determined and stored as an envelope of the series of periodic sensor signal curves, which fluctuate statistically, taking into account an additional offset or a factor.

[0043] Preferably, the detector algorithm is designed to perform and display a trend analysis of the sensor signal or a course of the sensor signal during the determination of the first and second states.

[0044] Preferably, the detector algorithm is also configured to perform a trend analysis of the sensor signal or the profile of the sensor signal during the determination of the first and second states, and to repeatedly redetermine and store the predetermined setpoint value or the setpoint curve and the predetermined tolerance range. The trend of the sensor signal or the profile of the sensor signal is preferably displayed on the display unit 8.

[0045] Preferably, the alarm signal, which can be acoustic and / or optical and / or functional, is emitted at a predetermined distance of the sensor signal from the setpoint or the setpoint curve or when the second state is reached.

[0046] Preferably, the detector algorithm is designed to detect a periodicity in the sensor signal and to assign the sensor signal to the tolerance range in a timely manner. In other words, the sensor signal must be compared with a typical sensor signal profile in order to correctly assign the tolerance range to the sensor signal. Fig. Figure 3 shows a periodic sensor signal S1 and the tolerance range formed between the upper threshold curve L1 and the lower threshold curve L2. The detector algorithm must now be designed to compare the periodically fluctuating sensor signal S1 with the correct tolerance range. In the example shown, a period of the sensor signal S1 extends from t0 to t1, with the sensor signal S1 then continuing to run in a similar periodic manner along the time axis t after t1. A periodic sensor signal S1 is usually generated when the sheet metal strip has equidistant unevenness or holes or cutouts therein. For example, cutouts are punched into the sheet metal strip at evenly spaced intervals by a preceding punch, so that the sheet metal strip is then easier to bend or roll form in an area with the cutout; as a result, lower forces then occur in this area, which can be measured by the at least one sensor.If, for example, a section is missing, this can be detected by the detector algorithm based on the deviation in the sensor signal.

[0047] Preferably, the roll forming system also includes a computing unit or is connected to the computing unit, which is equipped with a model calculation module for a Finite Element Analysis (FEA) model calculation. Using the FEA, the predetermined setpoint and / or setpoint curve can be determined from data of the sheet metal profile before and during processing by the respective sensor roll forming stand 5 and transmitted to the microprocessor unit 6. The computing unit can be part of the roll forming system or remote from it and can be connected to the roll forming system, for example, via a LAN or Internet connection. Fig. 4 shows an example of a profile of the sheet metal strip 9 with arrow-like force vectors, which have been calculated by an FEA analysis in a forming area and can be used as the target values ​​or target value curves.

[0048] Preferably, the microprocessor unit 6 comprises a second detector algorithm configured with pattern recognition, which compares the at least one sensor signal or the profile of the sensor signal with a recognition pattern or recognition parameters and detects a further state when a predetermined correlation value is exceeded. The further state can, for example, be a state of the sensor roll-forming stand 5 with bearing damage, a change in sheet thickness, an adjustment, or a distortion of the sensor roll-forming stand 5 relative to an adjacent roll-forming stand.

[0049] Preferably, the second detector algorithm comprises a learning unit or a corresponding learning algorithm which, when the further state is present in response to an operator signal, determines a change in the at least one sensor signal or in the course of the sensor signal from the first state to the further state and stores it as the recognition pattern or the recognition parameters for the second detector algorithm.

[0050] Likewise, the detector algorithm can also comprise a learning unit or a corresponding learning algorithm which, upon the presence of the second state, which is communicated to the detector algorithm by a corresponding operator signal, determines a change in the at least one sensor signal or the progression of the sensor signal from the first state to the second state and stores it as the recognition pattern or the recognition parameters for the detector algorithm. Preferably, the detector algorithm also has one or more operator signals as an input signal in order to be able to learn or assign one or more states accordingly based on the one or more operator signals. Cluster correlation algorithms, for example, are known for this purpose. Other learning algorithms known from the prior art are also conceivable for use in this context.

[0051] Preferably, the detection algorithm and / or the second detection algorithm is based on a neural network, a Kalmann filter, a prediction filter or another known pattern recognition algorithm.

[0052] Preferably, a means of the roll forming system is designed to store previously determined recognition patterns or the recognition parameters for the respective second or further state in a memory or a database and to make them accessible to the Da or the second detector algorithm.

[0053] Preferably, the means of the roll forming system which stores the determined recognition patterns or the recognition parameters for the respective further state in a memory or a database and makes them accessible to the second detector algorithm, upon renewed operator input of the second or further state, again determines the respective recognition patterns or recognition parameters taking into account the previously stored recognition patterns or recognition parameters and stores them.

[0054] Preferably, the sensor signal, the predetermined setpoint or the predetermined setpoint curve and the permissible tolerance range each comprise two or more signal components, which are each processed together.

[0055] The sensor signal preferably comprises the first force component, which is introduced by the sheet metal strip perpendicularly onto the first roller, the second force component, which is introduced by the sheet metal strip tangentially onto the first roller and transversely to the longitudinal direction of the sheet metal strip, and / or the third force component, which is introduced by the sheet metal strip tangentially in the longitudinal direction onto the first roller. The sensor signal preferably also comprises the fourth force component, which is introduced by the sheet metal strip perpendicularly onto the second roller, the fifth force component, which is introduced by the sheet metal strip tangentially onto the second roller and transversely to the longitudinal direction of the sheet metal strip, and / or the sixth force component, which is introduced by the sheet metal strip tangentially in the longitudinal direction onto the second roller.

[0056] Preferably, the sensor signal comprises a difference signal between the first force component introduced by the sheet metal strip perpendicularly onto the first roller and the fourth force component introduced by the sheet metal strip perpendicularly onto the second roller.

[0057] Preferably, the sensor signal also comprises at least one component of at least one adjacent sensor roll forming stand 5 or at least one other sensor in or on the roll forming system.

[0058] For example, the sensor signal may comprise at least one current component of at least one drive motor in the roll forming system.

[0059] Preferably, the sensor roll forming stand 5 is arranged between a first and a last roll forming stand.

[0060] A preferred method for roll forming or bending the sheet metal strip in the roll forming system with the plurality of roll forming stands arranged one behind the other, wherein at least one of the roll forming stands is designed as the above-described sensor roll forming stand 5 with the microprocessor unit 6 and display unit 8 connected thereto, comprises the following steps: a) feeding and passing the sheet metal strip through the roll forming machine; b) measuring at the respective sensor roll-forming stand 5 at least one force and / or one torque between the sheet metal strip and the respective sensor roll-forming stand 5, wherein the respective sensor signal belonging to the respective sensor roll-forming stand 5 or the respective signal component of the respective sensor signal is generated; c) Determining and storing the target sensor signal belonging to the respective sensor roll forming stand 5, which corresponds to the target value or the target value curve and is assigned to a respective desired process processing on the respective sensor roll forming stand 5; d) Determining and storing the respective tolerance range associated with the respective target sensor signal, in which the respective sensor signal statistically lies in the first state of the respective sensor roll forming stand 5; e) Repetitive measurement of the respective sensor signal and comparison with the respective tolerance range; and f) Determining and displaying the first state if the respective sensor signal lies within the respective tolerance range, and otherwise determining and displaying at least one second state of the respective sensor roll forming stand 5.

[0061] Preferably, the determination of the respective target sensor signal and / or the respective tolerance range is carried out statistically using the respective operator signal, through which the first and / or the at least one second state of the respective sensor signal of the respective sensor roll forming stand 5 is assigned by an operator. For example, each time the operator signal is activated, the respective sensor signal or a profile of the respective sensor signal is stored and, more preferably, averaged with a previously determined respective sensor signal or profile of the respective sensor signal.

[0062] Preferably, the respective target sensor signal and / or the respective tolerance range can also be determined using a finite element analysis, wherein at least one force and / or the torque between the sheet metal strip and the respective sensor roll-forming stand 5 is determined and the associated tolerance range is estimated. For clarity, the target sensor signal can also be a curve of the target sensor signal.

[0063] Preferably, the determination of the respective target sensor signal and / or the respective tolerance range is carried out continuously or periodically by a learning algorithm which expands the tolerance range for the detection of the first state for the sensor signal until at least the second state of the respective sensor roll forming stand 5 is input and assigned by an operator signal.

[0064] Preferably, after a first phase of determining the respective target sensor signal and / or the respective tolerance range, the respective target sensor signal and / or the respective tolerance range are then further determined continuously or periodically by the learning algorithm in the normal operating phase, which expands the tolerance range for detecting the first state for the sensor signal until the at least one second state of the respective sensor roll forming stand 5 is input and assigned by an operator signal.

[0065] Preferably, in response to an operator signal indicating a fault condition on the respective sensor frame 5, a pattern of the sensor signal or the profile of the sensor signal is determined and stored, wherein the pattern is significant for the fault condition. The sensor signal is then further analyzed for this fault condition using a corresponding detection or pattern recognition algorithm, and if a correlation value is exceeded, the fault condition is determined and displayed.

[0066] For the sake of clarity, the terms “top” and “bottom” are understood to mean relative locations in a vertical direction, as shown in the figures. For clarity, the sensor signal is a time-dependent sensor signal and can also be viewed as a sensor signal waveform. A roller is synonymous with a cylinder. Preferably, the first 3 or second shaft 4, not as in Fig.1, they can be supported not only at both ends, but also at one end only. The sensor roll-forming stand 5 or the other roll-forming stands can also be adjustable by actuators.

[0067] Further possible embodiments are described in the following claims. In particular, the various features of the above-described embodiments can also be combined with one another, provided they are not technically mutually exclusive.

[0068] The reference symbols mentioned in the claims serve only to improve comprehensibility and in no way limit the claims to the forms shown in the figures. List of reference symbols 1, 1' first sensor 2, 2' second sensor 3 first wave for the first roll 4 second shaft for the second roll 5 Sensor roll forming stand 6 Microprocessor unit 7 operating signal generators 8 Display unit or display 9 sheet metal strip S1 sensor signal L1 upper threshold curve L2 lower threshold curve t time t0 first time point t1 second time point x, y, z coordinates

Claims

[1] Roll forming plant for roll forming a sheet metal strip (9) with a plurality of roll forming stands arranged one behind the other, which form and / or guide the sheet metal strip (9) in sequence, comprising: a) at least one sensor roll forming stand (5) which is arranged in the plurality of roll forming stands arranged one behind the other and which forms, bends and / or guides the sheet metal strip (9); b) a first roller and a second roller connected to the respective sensor roll forming stand (5) for rotation about a respective rotation axis and held in position thereby, the first and second rollers being arranged to clamp the sheet metal strip (9) therebetween; c) at least one sensor (1, 2) arranged in the respective sensor roll-forming stand (5) to measure a force, a vibration or a torque exerted by the respective sensor roll-forming stand (5) on the sheet and generates a corresponding sensor signal (S1); characterized by d) a microprocessor unit (6) with a detector algorithm designed to compare the sensor signal (S1) with a corresponding predetermined setpoint value or a setpoint curve at a corresponding predetermined permissible tolerance range, wherein if the sensor signal (S1) is detected to be within the permissible tolerance range, a first state and otherwise at least a second state of the respective sensor roll forming stand (5) is detected; and e) a display unit (8) which displays the detected first and / or second state for the respective sensor roll forming stand (5), f) wherein the detector algorithm is designed to sample, store and statistically evaluate the sensor signal (S1) over a predetermined period of time in response to an operator signal in order to determine and store the setpoint value or the setpoint value curve and / or the permissible tolerance range. [2] Roll forming system according to claim 1, wherein the detector algorithm determines the predetermined target value over the predetermined period of time as an arithmetic mean, a median value or other mean value; and / or wherein the detector algorithm determines the predetermined permissible tolerance range over the predetermined period of time via a minima and maxima detection with an additional tolerance offset or factor offset; or wherein the detector algorithm determines the predetermined setpoint curve and the permissible tolerance range over the predetermined period of time in such a way that the detector algorithm detects a periodicity in the sensor signal (S1) with a known period detection algorithm and thus decomposes the sensor signal (S1) into a series of periodic sensor signal curves, wherein the period detection algorithm determines an averaged curve from the series of periodic sensor signal curves as the predetermined setpoint curve, and the permissible tolerance range is determined and stored as an envelope of the series of statistically fluctuating periodic sensor signal curves, taking into account an additional offset or a factor. [3] Roll forming system according to one of the preceding claims, wherein the detector algorithm is designed to perform and display a trend analysis of the sensor signal (S1) or a profile of the sensor signal (S1) during the determination of the first and second states; and / or wherein the detector algorithm is designed to perform a trend analysis of the sensor signal (S1) or the profile of the sensor signal (S1) during the determination of the first and second states, wherein the predetermined setpoint value or the setpoint value curve and the predetermined tolerance range are repeatedly redetermined and stored; and / or wherein a trend of the sensor signal (S1) or the course of the sensor signal (S1) is displayed on the display unit (8); and / or wherein an acoustic and / or optical and / or functional alarm signal is triggered at a predetermined distance of the sensor signal (S1) from the setpoint or the setpoint curve or when the second state is reached; and / or wherein the detector algorithm is designed to detect a periodicity in the sensor signal (S1) and to assign the sensor signal (S1) to the tolerance range in a timely manner in comparison with the tolerance range. [4] Roll forming system according to one of the preceding claims, further comprising a computing unit with a model calculation module for a finite element model calculation in order to determine the predetermined setpoint value and / or the setpoint curve from data of the sheet metal profile before and during processing by the respective sensor roll forming stand (5) and to transmit it to the microprocessor unit (6). [5] Roll forming system according to one of the preceding claims, wherein the microprocessor unit (6) comprises a second detector algorithm which is designed with a pattern recognition which compares the at least one sensor signal (S1) or the course of the sensor signal (S1) with a recognition pattern or recognition parameters and, when a predetermined correlation value is exceeded, recognises a further condition, such as preferably a bearing damage, a sheet thickness change, an adjustment or a distortion of the sensor roll forming stand (5) in relation to an adjacent roll forming stand or another change. [6] Roll forming system according to claim 5, wherein the second detector algorithm comprises a learning unit or a corresponding learning algorithm which, when the further state is present, determines a change in the at least one sensor signal (S1) or the course of the sensor signal (S1) from the first state to the further state in response to an operator signal and stores it as the recognition pattern or the recognition parameters for the second detector algorithm. [7] Roll forming plant according to claim 5 or 6, wherein the second detection algorithm is based on a neural network, a Kalmann filter, a prediction filter or another known pattern recognition algorithm. [8] Roll forming system according to one of the preceding claims 5 to 7, wherein a means of the roll forming system is designed to store previously determined recognition patterns or the recognition parameters for a respective further state in a memory or a database and to make them accessible to the second detector algorithm; and / or wherein the means of the roll forming system which stores the determined recognition patterns or the recognition parameters for the respective further state in a memory or a database and makes them accessible to the second detector algorithm, when the respective further state is entered again by the operator, again determines and stores the respective recognition patterns or recognition parameters taking into account the previously stored recognition patterns or recognition parameters. [9] Roll forming system according to one or more of the preceding claims, wherein the sensor signal (S1), the predetermined setpoint or the predetermined setpoint curve and the permissible tolerance range each comprise two or more signal components which are each processed together; wherein the sensor signal (S1) comprises a first force component which is introduced perpendicularly from the sheet metal strip (9) onto the first roller; and / or wherein the sensor signal (S1) comprises a second force component which is introduced by the sheet metal strip (9) tangentially onto the first roller and transversely to a longitudinal direction of the sheet metal strip (9); and / or wherein the sensor signal (S1) comprises a third force component which is introduced from the sheet metal strip (9) tangentially in the longitudinal direction onto the first roller; and / or wherein the sensor signal (S1) comprises a fourth force component which is introduced perpendicularly from the sheet metal strip (9) onto the second roller; and / or wherein the sensor signal (S1) comprises a fifth force component which is introduced by the sheet metal strip (9) tangentially onto the second roller and transversely to the longitudinal direction of the sheet metal strip (9); and / or wherein the sensor signal (S1) comprises a sixth force component which is introduced from the sheet metal strip (9) tangentially in the longitudinal direction onto the second roller. [10] Roll forming system according to one or more of the preceding claims, wherein the sensor signal (S1) comprises a difference signal between the first force component introduced by the sheet metal strip (9) perpendicularly onto the first roll and the fourth force component introduced by the sheet metal strip (9) perpendicularly onto the second roll; and / or wherein the sensor signal (S1) comprises at least one signal component of at least one adjacent sensor roll forming stand (5); and / or wherein the sensor signal (S1) comprises at least one signal component of a current of at least one drive motor. [11] Roll forming plant according to one or more of the preceding claims, wherein the sensor roll forming stand (5) is arranged between a first and a last roll forming stand. [12] Method for roll forming or bending a sheet metal strip (9) in a roll forming plant with a plurality of roll forming stands arranged one behind the other, wherein at least one of the roll forming stands is designed as a sensor roll forming stand (5) with a microprocessor unit (6) and display unit (8) connected thereto, according to one of the preceding claims 1 to 12, comprising the following steps: a) feeding and passing the sheet metal strip (9) through the roll forming system; b) measuring at the respective sensor roll-forming stand (5) at least one force and / or one torque between the sheet metal strip (9) and the respective sensor roll-forming stand (5), wherein a respective sensor signal (S1) or at least one respective signal component of the sensor signal (S1) is generated; c) determining and storing a respective target sensor signal which is assigned to a respective desired process on the respective sensor roll forming stand (5); d) determining and storing a respective tolerance range associated with the respective target sensor signal, in which the respective sensor signal (S1) statistically lies in a first state of the respective sensor roll forming stand (5); e) Repetitive measurement of the respective sensor signal (S1) and comparison with the respective tolerance range; and f) Determining and displaying the first state if the respective sensor signal (S1) lies within the respective tolerance range, and otherwise determining and displaying at least one second state of the respective sensor roll forming stand (5) characterized by , that g) the determination of the target sensor signal and the tolerance range is carried out in such a way that the sensor signal (S1) is sampled over a predetermined period of time in response to an operator signal, stored and statistically evaluated by a detector algorithm in order to determine and store the target sensor signal and / or the permissible tolerance range. [13] Method according to claim 12, wherein the detector algorithm for determining the respective target sensor signal and / or the respective tolerance range comprises and is carried out by a finite element analysis which determines the at least one force and / or the torque between the sheet metal strip (9) and the respective sensor roll forming stand (5), wherein the associated tolerance range is determined in an estimated manner. [14] Method according to claim 12, wherein the detector algorithm comprises a learning algorithm which determines the respective target sensor signal and / or the respective tolerance range continuously or periodically and which expands the tolerance range for a detection of the first state for the sensor signal (S1) until at least a second state of the respective sensor roll forming stand (5) is input and assigned by a further operator signal. [15] Method according to one of the preceding claims 12 to 13, wherein after a first phase of determining the respective target sensor signal and / or the respective tolerance range in the subsequent operating phase, the respective target sensor signal and / or the respective tolerance range are further determined continuously or periodically by the learning algorithm, which spreads the tolerance range for a detection of the first state for the sensor signal (S1) until the at least one second state of the respective sensor roll forming stand (5) is input and assigned by a further operator signal. [16] Method according to one of the preceding claims 12 to 15, wherein, in response to a further operator signal relating to a fault condition on the respective sensor frame (5), a pattern of the sensor signal (S1) or of the course of the sensor signal (S1) is determined and stored which is significant for the fault condition, whereupon the sensor signal (S1) is subsequently further examined for this fault condition by a corresponding detection or pattern recognition algorithm and, if a correlation value is exceeded, the fault condition is determined and displayed.

Images