MXPA00003124A - Method of and apparatus for deterministically obtaining measurements of a process control device parameter while a process is operating on-line - Google Patents

Abstract

A diagnostic test unit for deterministically measuring one or more parameters, such as dead band, dead time, response time, gain, or overshoot, of a process control device that is connected in a process control loop during operation of a process includes a switch controller, a signal generator, a switch, a response accumulator mechanism, and an analyzer mechanism. The swtich controller monitors a process signal during operation of the process to determine whether the process signal is substantially stable. In the event that the process signal is substantially stable, the switch replaces a control signal with a diagnostic test signal generated by the signal generator. The response accumulator is in communication with the process control loop to obtain an indication of the response of the process control device to the diagnostic test signal. The analyzer unit then determines the device parameter from the test signal and the response indication.

Description

METHOD AND APPARATUS FOR OBTAINING MEASUREMENTS OF A PAR METRO OF A PROCESS CONTROL DEVICE IN A MANNER DETERMINISTICA, WHILE A PROCESS IS OPERATING ONLINE The present invention relates generally to the diagnosis of the process control system and, more particularly, to a method of and an apparatus for obtaining measurements of one or more parameters of a process control device in a deterministic manner. which connects within a process environment in operation.

BACKGROUND OF THE ART Large commercial manufacturers and refining processes typically use a process controller to control the operation of one or more process control devices, such as control valves, which in turn control one or more process variables. such as, fluid flow, temperature, or pressure within the process. Generally, a process control valve has an actuator that is controlled by a positioner that moves an associated control element, such as a valve plug, a damper or some other changeable opening element, in response to the signal of control generated by the process controller. The control element of a control valve may, for example, move in response to the changing pressure of the fluid in a spring-biased diaphragm or a piston head, or in response to the rotation of an arrow, each of which it may be caused by a change in the control signal. In a standard valve mechanism, a control signal with a magnitude that varies in the range of 4 to 20 mA (milli-amperes), causes a positioner to alter the amount of fluid and therefore the fluid pressure. , inside a pressure chamber in proportion to the magnitude of the control signal. Changing the fluid pressure in the pressure chamber causes a diaphragm to move against a biased spring, which, in turn, causes the movement of a valve plug coupled to the diaphragm. The process control devices usually develop or produce a regeneration signal indicative of the response of the device to the control signal and supply this regeneration signal (or response indication) to the process control device for use in controlling a control. process. For example, control valves typically produce a regeneration signal that indicates the position (ie travel) of a valve plug or other moving element of the valve. Although control valves can use these regeneration signals to perform functions within a process control cycle, it has been found that poor cycle performance can still be caused by poor valve operating conditions. of control. In many cases, the problems associated with the individual process control devices can not be detuned from the control cycle by the process controller and, as a result, the poor performance control cycles are put into manual operation or are detuned until the point where these are effectively in manual operation. In this case, the processes associated with these control cycles require the constant supervision of one or more experienced human operators, which is not desirable. The poor performance of a control cycle can usually be overcome by monitoring the operational condition or "health" of connected process control devices within a process cycle, and by repairing or replacing control devices of process that operate poorly. The health of a process control device can be determined by measuring one or more parameters associated with the process control device and determining whether one or more parameters are outside an acceptable range. A parameter of the process control device can be used to determine, and the foregoing is indicative of, that the health of a process control device is dead band. Generally speaking, in the instrumentation of the process, the dead band is the range through which the input signal can be varied, by reversing the direction, without initiating a visible change in an output signal. The dead band, which can be caused by the physical play between mechanically interconnected components, friction, and / or any other physical phenomenon, is best observed when a control signal causes a counter-movement in the direction of movement of a moving element of a process control device. During this counter-march, the control signal is subject to a discrete amount of change (dead band) before the mobile element of the process control device actually shows movement in the new direction. Put another way, the difference between the value of the control signal in motion of the element of the process control device in a first direction that occurred last and the value of the control signal in motion of the element of the control device of process occurred first in a second and different direction is a measure of the dead band of the process control device. With reference to Figure 1, approximate estimates of the deadband have been obtained by applying a blocked sinusoidal signal to a process control device. The blocked sinusoidal signal includes periods of alternating steps of equal magnitude increasing in amplitude from period to period, such as 1 percent, 2 percent, 5 percent and so on. Once the movement occurs * # & * to a valve element or the process variable occurs after the counter-direction, the amplitude of the step (duplicate) provides an upper limit on the dead band. The lower limit is given by the amplitude of the steps in the previous period. Other device parameters that can be used to determine the health of a process control device are dead time, response time, gain, and overshoot. Downtime is associated with, and can be considered as a measure of the amount of time that the process control device requires to actually start moving a moving element in response to a change in a control signal. The response time is the amount of time that the mobile element of a process control device needs to reach a certain percentage, for example 63 percent, of its final value in response to a change in a control signal. The gain of a process control device is indicative of the amount of amplification caused by a change in the control signal. The gain can be expressed as the ratio of the relative change in a valve travel to the relative change in the control signal. The overshoot of a process control device indicates how much a valve travels beyond its eventual fixed-state position. If the dead band, downtime, response time, or other parameter (s) of process control devices of a process control device increases a significant amount over its nominal values, it may be necessary to repair or replace it e process control device to establish adequate control within the process control cycle. However, it is usually not very easy to measure the parameters of the process control device, such as dead band, dead time, response time, gain, and overshoot to monitor the health of process control devices. that are working when those devices are connected online within a control cycle. In the past, operators have had to eliminate a process control device from a control cycle to perform a stepped test of the device or, alternatively, control cycles have been provided with bypass valves and redundant process control devices for enable the deviation of a particular process control device to test this device through the same while the process is in operation. Otherwise, operators have had to introduce significant disturbances within the process operation, or wait until the process stops or is in a programmed operation close to test the individual process control devices within the process. Each of these options is time consuming, expensive and potentially destructive to the process, while continuing to provide intermittent measurements of the parameters of the individual process control devices that are required to determine the operating condition of these control devices.

COMPENDIUM OF THE INVENTION The present invention is directed to a method of and an apparatus for deterministically measuring one or more device parameters, such as deadband, dead time, response time, gain, or overshoot, of a control device. of process connected within a process while the process is operating (that is, while the process is online). The operation of the method and apparatus of the present invention allows an operator of a process to monitor the health or operating condition of a process control device within a process without having to remove the process control device from the control cycle, without having to bypass the process control device in the control cycle and without having to turn off the process or interfere with the process in any other meaningful way. For this purpose, the diagnostic test according to the present invention is preceded by a determination that the impact on the process will probably be minimal.

In accordance with one aspect of the present invention, a diagnostic test unit determines a device parameter associated with a process control device that is placed within an operating process. The diagnostic test unit includes a switch controller that monitors a signal from the process during the operation of the process. In addition, the diagnostic test unit includes a signal generator that produces a diagnostic test signal and a switch that responds to the switch controller and is operated to replace a control signal for the process control device with the signal of diagnostic test during the operation of the process. The apparatus further includes a mechanism for obtaining an indication of the response of the process control device to the diagnostic test signal and an analyzer unit that determines the device parameter of the diagnostic test signal and the response indication. The process control device can be a control valve having a movable valve element. In that case, the response indication is preferably a position signal that generates a position sensor in communication with the valve element, and the position signal is representative of the position of the valve element. The position signal can also serve as the process signal that the switch controller monitors.

In accordance with another aspect of the present invention, a method for determining a device parameter associated with a process control device while the process control device is placed within a process in operation includes the steps of monitoring a signal process during the operation of the process and determine if the process signal is substantially stable. The method further includes the steps for generating a diagnostic test signal and for replacing the control signal with the diagnostic test signal during the operation of the process if the process signal is substantially stable. Then, an indication of the response of the process control device to the diagnostic test signal is received and the device parameter is determined therefrom.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 comprises a graph of a deterministic test signal of the prior art which the diagnostic test unit of the present invention can use to measure the dead band of a process control device; Figure 2 comprises a block diagram of a process control device that is placed in a control cycle wherein the process control device comprises a diagnostic test unit in accordance with the present invention that is placed inside a control cycle; Figure 3 comprises a graph of a diagnostic test signal and a response thereto which is used to measure the dead band of a process control device according to the present invention; and Figures 4A-4D comprise graphs of deterministic test signals that are used to measure the parameters of a process control device in accordance with the present invention.

DETAILED DESCRIPTION Referring to Figure 2, a single input process control cycle, a single output 10 includes a process controller 11 that for example sends a 4 to 20 mA control signal to a control device. Process 13. The process control device 13 is illustrated as a control valve device that includes a switch 14, a printed circuit board (PWB) 15, a current-to-pressure transducer (I / P) 16, a relay 17, and an actuator / valve assembly 18. During normal operation, the control signal from controller 11 to PWB 15 is supplied through the switch 14. A position sensor 19 supplies a regeneration signal to the PWB 15 which indicates the movement and position of a moving valve element (not shown) that is placed within the actuator / valve assembly 18. The position of the the valve controls a variable process within a process 20. PWB 15 executes a control algorithm in accordance with input control and regeneration signals to develop a signal for the pressure-to-pressure transducer 16, which, in turn, develops a corresponding pressure signal. The relay 17 amplifies the pressure signal, which may comprise a disk valve with vertical movement or, more commonly, any pneumatic amplifier. The amplified pressure signal pneumatically controls an actuator (not shown) within the actuator / valve assembly 18 to move the valve member to the desired position. Both the current-to-pressure transducer 16 and the relay 17 develop the respective pressure signals using a pressure source 27 coupled to the process control device 13. The position sensor 19 can comprise any motion or position measurement device. desired, which includes, for example, a potentiometer, a linear variable differential transformer (LVDT), a rotating variable differential transformer (RVDT), a Hall effect motion sensor , a restrictive magnet motion sensor or a variable capacitor movement sensor. If desired, the process control device 13 may include other types of mechanisms or valve elements in place of or in addition to those illustrated in Figure 1, including, for example, a separate pneumatic positioner and an I / unit. P. Furthermore, it should be understood that the process control device 13 can be any other type of device, such as a damper or fan, that controls the variable of a process in any other way that is desired or known. As illustrated in Figure 1, a transmitter 22 measures the process variable of the process 20 and transmits an indication of the process variable that was measured to a splice of sum 24. The splice of sums 24 compares the measured value of the variable of the process (converted to a normalized percentage) to an adjustment point to produce an error signal that indicates the difference between them and supplies this error signal to the process controller 11. The set point, which can be generated by a user, operator or other controller (not shown), typically normalizes to be between 0 and 100 percent and indicates the desired value of the process variable. During the normal operation of the process 20, the process controller 11 uses the error signal to generate the control signal in accordance with any desired technique and sends the control signal to the process control device 13 to control the variable of the process. In accordance with the present invention, a diagnostic test routine, which may be pre-programmed, is implemented to test the process control device 13 under normal process operating conditions, that is, while the process 20 is in progress. line. In the event that it has been determined that the implementation of the diagnostic test routine would have minimal impact on the process 20, the diagnostic test routine disconnects the process control device 13 from the controller 11 and forces the control device of process 13 that performs a set of deterministic or determined operations previously designed to impact the process only in a minimal way, if at all. The routine also receives, measures, and collects the information that is indicative of the response of the process control device 13 to the diagnostic test signal and then again connects the process control device 13 to the controller 1.1, all of this within a limited amount of time to avoid altering the normal operation of the process 20 in a substantial manner. The information that was collected can be used to calculate the parameters of the process control device such as dead band, dead time, response time, gain, and overshoot, with the interest of determining the operation condition of the process control device 13. Since the diagnostic routine is implemented while process 20 is online, the "health" or operating condition of the process control device 13 is determined without having to isolate or bypass the process control device 13 and / or turn off the process 20. When a diagnostic test of the process control device 13 in accordance with the present invention, a switch contributor 25 generates (or modifies) a switch signal to articulate the switch 14 of a first position (or state) in which the switch 14 supplies the signal of control of the controller 11 to the PWB 15 to a second position (or state) in which the switch 14 disconnects the controller 11 and, therefore, the control signal, of the PWB 15 and connects the PWB 15 to the output of a generator signal 26, which produces a deterministic (ie, known or previously determined) diagnostic test signal. As discussed above, although the switch controller 25 replaces the control signal with the diagnostic test signal, the process 20 remains online during the diagnostic test routine. Both the switch controller 25 and the signal generator 26 are part of a diagnostic test unit 27, which can be internal to the process control device 13 as shown in Figure 2 or, alternatively, can be a external test apparatus coupled to the process control device 13. Similarly, the switch 14 may be internal or external to the process control device 13. The diagnostic test unit 27 further includes a response accumulator 28 which collects or receives one or more indications of the response of the process control device 13 to the diagnostic test signal. The response accumulator 28 may comprise a memory or storage device that stores the indications of the response to supply the data representative of the indications of the response to an analyzing unit 29. The analyzing unit 29, which may also receive the data representative of The diagnostic test signal of the signal generator 26 analyzes the test signal and the response indication data to determine one or more desired parameters of the process control device. As illustrated in Figure 1, the response accumulator 28 can receive a response indication that is indicative of the movement or position of the valve (valve stroke) of the position sensor 19. Alternatively, or in addition, the response accumulator 28 can receive, as a response indication, the command signal of the actuator that developed the relay 17 (by means of a pressure sensor 36), and / or any other signal that specifies or relates to the control of the process control 13 such as the output of the transmitter 22 which indicates the value of the process variable. It should be noted that other types of process control devices may have other signals or phenomena associated therewith which may also indicate a response to a diagnostic test signal. In this way, in general, the response accumulator 28 can pick up or receive any signal or phenomenon indicating the movement or operation of the process control device 13 in response to a change in the diagnostic test signal. However, some response indications, such as the position of the valve, can provide more accurate estimates of device parameters by avoiding sources of noise (eg, process noise) that is not associated with the control device. of process 13 that is being tested. On the other hand, the response indications that are removed later, such as the process variable, may exhibit delays that are not related to health or the performance of the process control device 13 that still needs to be considered. Therefore, certain indications of response (or combinations thereof), may be preferable in certain situations. For the purpose of determining whether to initiate a diagnostic test, the switch controller 25 monitors a signal that is associated with the process 20 while the process 20 is online to determine whether the process variable or the process control device 13 is substantially stable. The process signal that was monitored by the response accumulator 28 can be received or directly picked up by the switch controller 25. The signal of the process can be the control signal, the position of the valve element, the process variable, or any other variable or signal that provides an indication of the extent to which the process control device 13, the process variable, and / or the set point are quiescent. If it is found that the set point, control signal or other variable is fluctuating, or fluctuates to a large extent, the diagnostic test routine will not be started. More than one variable can be monitored within the process cycle 10 by means of the switch controller 25 with the interest of improving the accuracy of the determination and with the interest of minimizing the effects of the implementation of the diagnostic test routine while process 20 is online. The stability of the monitored process signal (s) is indicative of the stability of the process variable, the process control device 13, or the set point. A substantially stable or immobile process signal may still fluctuate to some degree despite the process variable, the process control device 13, or the set point being stable. Therefore, the "substantially stable" determination should at least allow the inherent noise in the signal. A diagnostic test routine should not be implemented if the process signal is changing to the extent that it is indicative of significant changes to process 20 or the process variable. As a result, if the monitored process signals are considered to be substantially stable or immobile, it depends, to a great degree, on the signal (s) that are used as the monitored process signal. For example, a diagnostic test routine can still be implemented when a control signal is fluctuating up to and beyond 5 percent in certain process control cycles, while a routine could be harmful to process 20 if the variable The process is varying as little as 1 percent in other process control cycles. These differences can be the result of the nature of the process and / or the process variable, or that flow from a large differential in the dead bands between different process control devices. 13. Depending on the health and / or quality of the process. actuator / valve assembly 18, a large dead band can allow the operation of diagnostic test routines with control signal variations as large as 10 percent. In addition, a low sensitivity of process 20 to the process variable may promote that the permitted variation is even greater. The diagnostic test unit 27, and any component thereof, including the switch controller 25, the signal generator 26, the response accumulator 28 and / or the analyzer unit 29, can be implemented in the hardware, software and firmware, or any combination thereof. If implemented in the software, the components of the diagnostic test unit 27 can be stored in any memory device, such as a floppy disk, hard disk, CD-ROM, RAM, ROM, EEPROM or any other known storage medium. by those skilled in the art, and, if desired, can be supplied from a remote location by any means of communication, such as transmission by telephone lines, the Internet, an Ethernet, or any other type of communication network known to those skilled in the art. Similarly, switch 14 can be implemented in hardware, software, firmware, or any combination thereof. If desired, the analyzing unit 29 can compare the parameters of the determined process control device with one or more stored values to determine if the parameters that were measured are acceptable or are within one or more specified ranges. If the parameters of the process control device are not within one or more specified ranges, the analyzing unit 29 or another component of the diagnostic test unit 27 alerts the user by means of the visual display 38 comprising, for example, a CRT screen, a printer, a speech generator, an alarm, or any other communication device that is desired, that the process control device 13 needs to be repaired or replaced. Also, if desired, the analyzing unit 29 can provide the user with a list of the device parameters that were measured by means of the visual display 38. The diagnostic test signal produced by the signal generator 26 can assume any shape that it wishes and that allows the measurement of a device parameter of control of process. However, many waveforms of the diagnostic test signal that are described later herein and illustrated in Figures 1, 3, and 4A-4D can be used to obtain parameter measurements in a manner that minimizes the potential for harmful interference with the process 20. In general, the waveforms of the diagnostic test signal are plotted on a time axis to show the amplitude of the signal from the zero reference point that can constitute any DC level , including zero. For example, the amplitude values plotted in Figures 4A-4C may be representative of the amount that the amplitude of the diagnostic test deviates from the command signal amplitude of 4-20 mA when the command signal is turned off. process control device 13. Some of the diagnostic test signals are also shown with a frame of an exemplary response indication showing the movement of the response indication related to zero, which may represent a position or a value of signal at the beginning of the test routine. Referring now to Figure 1, a first diagnostic test signal is useful in the process of measuring the parameters of the process control device (particularly of the dead band) comprises a sinusoidal pulse signal having a sequence of steps that are configured in a multiplicity of periods. Each period (for example, from time T1 to time T4) can include a pair of alternating pulses which, more particularly, include a positive pass (for example, at time T -, _) a return-to-zero (for example at time T2), a negative step of the same magnitude (for example at time T3) and another return-to-zero. Preferably, the magnitude of the pulses increases during successive periods. To measure the deadband using the pulse sinusoidal signal of Figure 1, the signal generator 26 first supplies one or more periods of the alternating pulses until the movement of, for example, the valve element is detected by the sensor of position 19 during the impulses both positive and negative of any particular period. The absolute difference between the amplitudes of the period (which is usually expressed as a percentage of the stretch) during which the movement of the valve element occurs first in response to both positive and negative impulses, is a measure of the dead band. Of course, this measurement really overestimates the dead band. The succession from the low amplitudes driven (eg ax) to the higher pulse amplitudes (eg, a4) provides the lower and upper limits for the deadband. Another routine to determine the dead band can use the movement detected in multiple consecutive periods. For example, if the movement is detected first in a first direction (for example, due to the negative impulse) of a period and is detected first in a second direction due to the positive pulse of a later period, the difference between the negative pulse amplitude and the positive pulse amplitude (of the later period) can be used as a measure of the dead band of the device. As will be understood, the sinusoidal pulse signal of Figure 1 allows the process control device 13 to be bi-directionally tested around a given operating point. The sinusoidal pulse signal can also be used as an initialization or pretest sequence to ensure that the process control device 13 is against a bank of the deadband. The pulse amplitudes may increase linearly (as shown in Figure 1), non-linearly, or in any other way that is desired. In addition, the frequency of the diagnostic test signal must remain low enough to ensure that the valve 18 (or any other moving element) has achieved a fixed state between each of the steps. The frequency of the diagnostic test signal will generally depend on the particular process control device 13 being tested, but, in general, it can be as low as, for example, 0.2 Hz to 20 Hz. Referring now to Figure 3, a second diagnostic test signal that is useful for measuring the parameters of the process control device (particularly the dead band) comprises a ramp-passage signal having an initialization phase, a phase test, and a phase after the test. At time T0, the switch controller 25 initiates the diagnostic test by means of articulating the switch 14 and the signal generator 26 starts the initialization phase. In the initialization phase, the diagnostic test signal increases (or decreases) in either linear or variable speed steps until the movement of the valve element or other process variable is detected first (for example up to time T ?) The movement of the valve element (ie the change in the response indication) is also shown as an amplitude relative to zero, which can represent any initial position or response indication value. Then the diagnostic test signal remains constant while the valve element or the process variable reaches a fixed status value, new (for example, position P). The test phase starts at time T2 and the direction of the diagnostic test signal is reversed. Then, the test signal decreases (or increases) either in a linear or variable speed manner until the movement of the valve mechanism or the process variable is detected a second time (for example at time T3). After this second movement, the deadband can be estimated as the absolute difference in the amplitude of the test signal between times T2 and T3. Once the dead band has been estimated, the diagnostic test signal can return to zero at an accelerated speed in the subsequent test phase. For example, as shown in Figure 3 after time T3, the steps progress at a rate twice as fast as those during the test phase. Once the amplitude of the diagnostic test signal approaches zero, the signal generator 26 can return the diagnostic test signal to zero at a normal speed if the double speed steps could overshoot at zero. At time T4, the diagnostic test is complete and the switch controller 25 can articulate the switch 14 to return the control of the process control device 13 to the controller 11 or to proceed with a different test routine. Excessive deviation of the amplitude of the control signal (ie large deviations from zero) can result in harmful interference to the process 20. Accordingly, it is preferred that the diagnostic test signal deviate from zero only to the degree that is necessary. This activity is accomplished in the diagnostic test routine shown in Figure 3 through the use of very small steps (eg, as low as .25 percent interval) throughout the entire test routine. In this way, the edges of the dead band are reached and only slightly exceeded in such a way that the position (or value) P differs very little from the initial position (or value) that the zero represents. With such small test signal steps, the sensitivity to detect the movement of the valve (or a change in the response indication) is of paramount importance. For this purpose, some signals may be preferable to others for use as the response indication. For example, monitoring the movement of the valve element by means of the position sensor 19 can provide a more accurate estimate of the dead band of the actuator / valve assembly 18 than the estimate that was provided by monitoring the variable of the valve member. process. Movement monitoring of the valve element can also be preferred because movement can be detected before the process variable has even been changed, thereby minimizing the effect of the diagnostic test routine in the process 20 if the Valve element can be quickly returned to its initial position. As is evident from the above comments and Figures 1 and 3, the diagnostic test unit 27 and, in particular, the signal generator 26 have the ability to supply a multitude of different diagnostic test signals. In addition, each diagnostic test routine may comprise one or more different diagnostic test signals to determine multiple estimates of one or more device parameters. It is preferred that multiple estimates of each device parameter be obtained because the estimates of the device parameter may vary depending on whether the process control device 13 moves through the dead band of the process control device 13. In view of this potential to make an error, or at least, a variation, a diagnostic test signal may also include an initialization waveform that begins by moving the process control device 13 to a edge of the deadband. If the dead band has already been estimated previously, then the signal generator 26 supplies a diagnostic test signal comprising a step having an amplitude equal to the dead band plus an amount corresponding to, for example, one percent of the range of movement of the valve element. This step can be positive or negative and can be started anywhere inside or on the edge of the dead band. In this way, the response of the step must always be long enough to exceed the dead band. Then the response accumulator 28 and / or the analyzer unit 29 can measure (1) the time until the valve 18 moves to determine the dead time, (2) the time that the valve 18 is taken to move a previously determined percentage. (eg, 63 percent) of the complete change in the position of the valve to determine the response time, (3) the amount of overshoot of the valve 18 to the desired fixed-state value before it finally reaches its current value of steady state, and (4) the gain, if any, between the current and desired fixed state values as a result of a change in the diagnostic test signal. By making the response time test of the process control device 13, the percentage previously determined can be specified, for example, by means of an operator, a control cycle designer, or a component of the diagnostic test unit 27. As evidenced, diagnostic test unit 27 includes a stopwatch or other timing device for measuring dead time or response time. It should also be noted that, for the measurements of dead time and response time (as well as any other measurement), the diagnostic test unit 27 can only wait for a predetermined period of time (for example, three to four seconds). ) before terminating the diagnostic test and generating an error signal to alert the operator or the user by means of the visual display 38 (Figure 2) that the valve 18 has a malfunction. The time limit is also useful to avoid harmful interference of the diagnostic test routine with the process 20. Of course, the amount of time the diagnostic test unit 27 may be connected to the input of the process control device 13 will be variable according to the volatility of the control signal, as well as other characteristics of the control cycle 10. In a preferred embodiment of the present invention, a diagnostic test routine includes four secondary routines that are implemented to determine four measurements of the dead time, response time, gain, and overshoot. Each of the secondary routines, which can be operated independently or in combination with the other secondary routines, assumes that an estimate of the deadband has already been previously obtained by, and / or is stored in, the diagnostic test unit 27 The four secondary routines are referred to herein as an Up / Up test, an Up / Down test, a Down / Down test, and a Down test. / Up (Down / Up). With reference to Figure 4A, the test Up / Above uses a diagnostic test signal comprising an initialization phase, a test phase, and a post-test phase and is intended to determine different parameters of the device when the valve is both on the upper edge of the deadband and moving upwards. As with the deadband test shown in Figure 3, the diagnostic test routine starts at time T0 at which time the diagnostic test signal enters the initialization phase. In the initialization phase, the diagnostic test signal increases in linear or variable speed steps (i.e., a ramp pass signal) until the movement of the valve element (or other process variable) is detected. ) (that is, until time Tx). At that point, the diagnostic test signal remains constant while the position of the valve element (or process variable) reaches a new, fixed e_state value, after which the test phase begins. During the test phase (at time T2), a single, positive step having an amplitude equal to the dead band plus an amount corresponding to, for example, one percent of the range of motion of the valve element is applied to the process control device 13. Then the analyzing unit 29 can determine the dead time, the response time, the overshoot, and the gain as set forth hereinabove. After the position of the valve element (or another process variable) arrives at a new, steady-state condition (ie at time T3), the diagnostic test enters the later phase of the test where the diagnostic test signal can return to zero. a way of step stepped at a normal or accelerated speed. The Up / Down test (Figure 4B) is similar to the Up / Up test with the exception that the diagnostic test signal comprises a single, negative step of equal magnitude. The Up / Down test measures the parameters of the device for the downward movement of the valve element through the dead band. The test Down / Down (Figure 4C) is, in turn, similar to the Up / Down test, but with a ramp-passage signal that has negative amplitude steps that moves the valve element to the lower edge of the deadband before it apply the signal of the single, negative step. The Down / Down test measures the parameters of the device for the downward movement of the valve element from the lower edge of the dead band. Finally, the Down / Up test (Figure 4D) uses a diagnostic test signal comprising a ramp-passage signal having negative steps and a single positive step, as described in relation to the Up / Up test. The Down / Up test measures the parameters of the device for the upward movement of the valve element through the dead band. Then, the estimates that are obtained from the parameters by these four tests, can be averaged or, combined in another way, by means of the analyzer unit 29 to determine a statistical measurement of the dead time, response time, gain, and overshoot of the process control device 13. Although the above description provides the means for calculating the dead band, dead time, response time, gain and overshoot of the process control device 13, the test signals of Diagnosis and response indications can also be used to calculate other parameters of the process control device. More particularly, any other parameter of the process control device can be obtained as long as it is obtained in a deterministic manner using a controlled diagnostic test signal and a measured response indication. Since the process is still online, the diagnostic test signal can be limited in both time and magnitude to ensure that it does not interfere with the process 20 to a degree that is harmful. In case the amplitude of the control signal, unlikely to change more than ten percent, the degree to which the amplitude of the diagnostic test signal can be biased from the amplitude of the control signal (when removes the control signal) can be very small, as is less than about five percent of the amplitude of the control signal. Fortunately, it is usually not necessary for the process control device 13 to be subjected to a full path sequence or test path sequence to determine the parameters of the device identified herein. In fact, in most cases, the device parameter can be estimated based on very small deviations in the amplitude of the control signal, such as less than about five percent and, preferably, less than about one percent. percent. In addition, the maximum time allowed for the diagnostic test will depend on the process control device 13 and the process involved, but will generally be less than about five seconds. During the diagnostic test routine, the diagnostic test unit 27, or some components of it (such as the switch controller 25, the signal generator 26, or the analyzer unit 29), may continue to monitor the control signal (or any other process variable or response indication) to ensure that the diagnostic test routine does not adversely affect the operation of the process 20. For example, the diagnostic test unit 27 can monitor to what extent the amplitude of the diagnostic test signal deviates from the amplitude of the control signal (which may change during the diagnostic test) . In the event that the deviation between the two amplitudes exceeds a predetermined amount, which can be determined by the user, an operator or the diagnostic test unit 27, the diagnostic test unit 27 can interrupt or terminate the routine of diagnostic test by directing the switch controller 25 to articulate the switch 14 to replace the diagnostic test signal with the control signal and / or by directing the signal generator 26 to generate a signal diagnostic test that has an amplitude that returns to the current value of the control signal. The degree to which the diagnostic test signal of the control signal can be diverted will depend on the characteristics of the particular process 20 and the process control device 13 in which the diagnostic test unit 27 is used. further preventing the operation of process 20 from being adversely affected, the diagnostic test routine according to the invention can operate in combination with a test apparatus that can measure the parameters of the process control device in a passive manner. For example, if the process 20 requires a modification of the control signal or the set point relatively frequently, the data representative of the frequent changes of the control signal and the response indications to the analyzer unit 29 can be supplied to the analyzer unit 29. the same so that the parameters of the device are calculated. In this case, the analyzer unit 29 should also include hardware, software and / or firmware to initially analyze the data to "determine when and if the data can be used to calculate the device parameters that are desired. of adjustment does not change frequently and, therefore, such a passive approach does not produce data, it may be necessary to test the process control device 13 in accordance with the present invention by means of disconnecting the controller 11 and applying a signal of Diagnostic test as set forth hereinabove It should be noted that the switch 14 can be placed downstream of the current-to-pressure transducer 16 before the pneumatic positioner (not shown) or, alternatively, current up as a component part of the controller 11. In the previous case, the diagnostic test signal is a pneumatic signal. You can use the same signal generator that produces the diagnostic test signal to develop the control signal. As a result, it should be understood that the switch 14 comprises a mechanism that simply replaces a signal with another signal by means of changing an output sequence. In still another embodiment, a user, operator or other controller (not shown) can generate the signal of the switch that articulates the switch 14 to manually force the initiation or termination of a diagnostic test routine. While the present invention has been described with reference to specific examples, which are intended to be illustrative only, and not limiting of the invention, it will be apparent to those of ordinary skill in the art that changes, additions or removals may be made. to the published modalities without departing from the spirit and scope of the invention.

Claims (30)

1. CLAIMS 1. A diagnostic test unit for determining a device parameter associated with a process control device that is placed within an operating process to receive a control signal, the diagnostic test unit comprising: a controller switch that monitors a process signal during the operation of the process; a signal generator that generates a diagnostic test signal: a switch that responds to the switch controller to replace the control signal with the diagnostic test signal during the operation of the process; elements for obtaining an indication of the response of the process control device to the diagnostic test signal; and an analyzing unit that determines the device parameter of the diagnostic test signal and the response indication. The diagnostic test unit according to claim 1, wherein the switch controller comprises elements for determining whether the process signal is substantially stable. 3. The diagnostic test unit according to claim 1, wherein the process signal is the control signal. 4. The diagnostic test unit according to claim 1, wherein the process signal is representative of a process variable controlled by the process control device. The diagnostic test unit according to claim 1, wherein the process control device comprises a valve having a movable valve member. 6. The diagnostic test unit according to claim 5, characterized in that it further comprises a position sensor that generates a position signal representative of the position of the valve member and where the position signal is the signal of the process and the position signal is obtained by means of the elements of obtaining as the response indication. The diagnostic test unit according to claim 1, wherein the device parameter is deadband. The diagnostic test unit according to claim 7, wherein the diagnostic test signal comprises a ramp-passage signal. 9. The diagnostic test unit according to claim 8, wherein the signal generator comprises elements for varying the ramp rate of the diagnostic test signal. The diagnostic test unit according to claim 1, wherein the process control device has a dead band associated therewith and the diagnostic test signal comprises a step signal having a greater amplitude than the of the dead band. The diagnostic test unit according to claim 10, wherein the diagnostic test signal comprises an initialization phase having a ramp-passage signal to bring the process control device to a edge of the band. dead. The diagnostic test unit according to claim 11, wherein the device parameter is selected from the group consisting of response time, dead time, gain, and overshoot. The diagnostic test unit according to claim 1, characterized in that it additionally comprises elements for alerting a user when the device parameter no longer satisfies a previously determined criterion. The diagnostic test unit according to claim 1, wherein the signal generator includes the elements to provide a multiplicity of different diagnostic test signals and wherein the analyzer unit includes the elements to determine an estimate of the parameter of device for each of the multiplicity of diagnostic test signals and elements to combine the estimates of device parameters to calculate the device parameter. 15. The diagnostic test unit according to claim 1, wherein the diagnostic test signal has an amplitude that deviates from an amplitude of the control signal by less than about five percent of a range of the control signal. 16. A control valve having a movable valve member positioned in a process in operation to receive a control signal, the control valve comprising: a switch controller that monitors a process signal during the operation of the process; a signal generator that generates a diagnostic test signal; a switch that responds to the switch controller to replace the control signal with the diagnostic test signal during the operation of the process; a position sensor in communication with the valve member to generate a position signal representative of the position of the valve member; and an analyzing unit that determines a device parameter of the diagnostic test signal and the position signal. 17. The control valve according to claim 16, wherein the switch controller comprises elements to determine if the process signal is substantially stable. 18. The control valve according to claim 16, wherein the process signal is the position signal. 19. A method for determining a device parameter associated with the process control device while the process control device is placed within an operating process to receive the control signal, the method comprising the steps of: monitoring a process signal during the operation of the process to determine whether the process signal is substantially stable; generate a diagnostic test signal; replace the control signal with the diagnostic test signal during the operation of the process if the signal of the process is substantially stable; receiving a response indication from the process control device to the diagnostic test signal; and determining the device parameter of the response indication. The method according to claim 19, characterized in that it further includes the step for replacing the diagnostic test signal with the control signal subsequent to the reception of the response indication. 21. The method according to claim 19, wherein the diagnostic test signal comprises a ramp-passage signal. 22. The method according to claim 21, wherein the ramp-passage signal comprises a first series of steps in a first direction and a second series of steps in a second direction, the device parameter is deadband, and the step to determine includes the steps of: detecting a first amplitude of the diagnostic test signal in which a first change occurs in the response indication during the first series of steps of the diagnostic test signal, detecting a second amplitude of the diagnostic test signal in which a second change occurs in the response indication during the second series of steps of the diagnostic test signal, and determine the dead band of the first amplitude and of the second amplitude. The method according to claim 19, wherein the process control device has a dead band and the diagnostic test signal comprises a first initialization phase having a ramp-passage signal to bring the control device of process to a first edge of the dead band followed by a first test phase having a step of an amplitude greater than that of the dead band. The method according to claim 23, wherein the diagnostic test signal further comprises a second initialization step having a ramp-passage signal to bring the process control device to a second edge of the bandpass. dead followed by a second test phase having a step of an amplitude greater than that of the dead band, t the receiving step comprising the steps to receive a first response indication in response to the first step and a second response indication in response to the additional step, and the determination step further comprises the step of combining respective device parameter values which is calculated from the first response indication and the second response indication to determine a statistical measurement of the device parameter. 25. The method of compliance with the claim 19, wherein the device parameter is selected from the group consisting of dead band, response time, gain, and overshoot. 26. The method according to claim 19, further comprising the steps of comparing the determined device parameter to a previously determined criterion, and alerting the user when the determined device parameter does not meet the previously determined criterion. 27. The method of compliance with the claim 19, where the process signal is the control signal. The method according to claim 27, characterized in that it further comprises the steps of comparing the amplitude of the control signal with the amplitude of the diagnostic test signal, and replacing the diagnostic test signal with the control signal if the amplitude of the diagnostic test signal deviates from the amplitude of the control signal by at least a predetermined amount. 29. The method of compliance with the claim 27, characterized in that it further comprises the step of replacing the diagnostic test signal with the control signal if the amplitude of the control signal changes by at least a predetermined amount. 30. The method of compliance with the claim 19, wherein the process control device comprises a valve having a movable valve member for controlling a process variable associated with the process and the response indication is a position signal representative of the position of the valve member.