AU2010236009A1

AU2010236009A1 - Method and apparatus for monitoring a process and/or a technical installation

Info

Publication number: AU2010236009A1
Application number: AU2010236009A
Authority: AU
Inventors: Frederik Blank; Markus Gauder; Stefan Lauxtermann; Werner A. Schmidt
Original assignee: ABB Technology AG
Current assignee: ABB Schweiz AG
Priority date: 2009-10-30
Filing date: 2010-10-25
Publication date: 2011-05-19
Anticipated expiration: 2030-10-25
Also published as: EP2320289A2; TW201126289A; DE102009051446B4; AU2010236009B2; DE102009051446A1; EP2320289A3

Description

AUSTRALIA Patents Act 1990 COMPLETE SPECIFICATION Standard Patent Applicant(s): ABB Technology AG Invention Title: Method and apparatus for monitoring a process and/or a technical installation The following statement is a full description of this invention, including the best method for performing it known to me/us: - 2 Method and apparatus for monitoring a process and/or a technical installation Description 5 The invention relates to a method and an apparatus for monitoring a process and/or a technical installation, wherein process values are determined, in particular measured, simulated or calculated, and are compared with respectively predetermined lower and/or upper limit 10 values, and if at least one lower limit value is undershot or at least one upper limit value is overshot then an alarm signal is given or produced. The monitoring and control of processes plays a critical role during the operation of installations and/or supply 15 systems. Apparatuses which detect and visualize data and values and which produce alarm signals are used for monitoring and control. Essentially, an apparatus monitors physical variables and values relating to them in order to provide an alarm 20 management system. As soon as one of the monitored values overshoots or undershoots at least one limit value associated with it, an alarm is produced, and this is signalled and visualized in an appropriate alarm list. The limit values are so-called alarm limits. Depending on the 25 alarm system design, the term upper/lower limit value in the following text includes one or more upper and one or more lower limit values, for example low, low low, high or high high, wherein the various alarm limit values and/or alarm ranges generally represent or denote the various 30 stages of the alarm state. Ideally, alarm signals are produced only when, for example, an installation operator or else a user should react, that is to say should carry out an action or else 2448964_1 (GHMattero)25/10/2010 -3 should take measures. Against this background, the number of alarm signals which actually occur should be kept as low as possible. Only alarm signals which are relevant or necessary for the respective process or the process 5 sequence and for which a measure must be taken should be triggered. In addition to alarm signals which are triggered by controllers in the field, that is to say in the case of .ocal process automation, and in particular at the 10 controller level, alarm signals also exist which are produced in a comprehensive monitoring system, specifically a control station, and in particular at control system level. These alarm signals are frequently also referred to as "soft alarms" or "smart alarms". 15 It is already known from the prior art for limit values to be set and to be permanently fixed statically at a constant value, for example for "soft alarms" for monitoring values in processes, in which case this value can be based on empirical values. 20 The statically set constant limit values and alarm limits formed from them are based on historical experiments, experiences or expert knowledge. These limit values are defined in the course of the engineering of an apparatus for monitoring processes. 25 External factors may, however, in this case lead to false alarm signals. One external factor, for example, may be a change in the outside temperature and/or the air humidity and/or some other measurable physical variable. By way of example, changes in consumption can be measured or 30 registered as an increased flow. In particular, the respective process values or components of an installation can be influenced and/or changed by fluctuations in external factors or by changing operating states, and a corresponding alarm may be triggered without a direct 2448964_1 (GWatter.)25/10/2010 - 4 fault or a malfunction having occurred in the installation and/or the supply system. In order to make it possible to ensure efficient monitoring and therefore handling and operation of a 5 technical installation and/or of a supply system as well as a process sequence which is as smooth as possible, it accordingly appears to be desirable and appropriate to reduce as far as possible the occurrence of such alarm signals which are not initiated by critical values of an 10 installation and/or a critical installation and/or process state, but are in fact caused by largely non-critical external factors and/or varying operating states. These alarm signals are also referred to in the following text as a false alarm. 15 The invention is therefore based on the object of designing and developing a method and an apparatus of the type mentioned initially such that an installation which is protected by alarms can be operated permanently with few disturbances, and the abovementioned false alarms are 20 very largely avoided. According to the invention, the abovementioned object is achieved by a method for efficient monitoring of a process and/or of a technical installation and/or of a supply system having the features of Patent Claim 1. Advantageous 25 refinements and developments of the method as well as an apparatus for efficient monitoring of a process and/or of a technical installation and/or of a supply system are specified in further claims and in the following description. 30 In the method according to the invention for efficient monitoring of a process and/or of a technical installation and/or of a supply system, at least one lower and/or at least one upper alarm limit is determined predictively and/or characteristic variables and/or measurement 35 variables as well as values relating to them in the 2448964_1 (G)10atters)25/10/2010 - 5 respective process and/or the respective technical installation and/or supply system are determined, are compared with predictively determined alarm limit values of at least one lower and/or at least one upper alarm 5 limit, and if a lower alarm limit is undershot or an upper alarm limit is overshot by a determined variable and/or a value determined in a manner relating to it, in particular a measured, simulated or calculated value, then an alarm signal is generated, wherein at least one lower and/or at 10 least one upper alarm limit is variable on a state basis and/or as a function of the environment, and is dynamically adaptable, in particular according to the process behaviour and/or installation behaviour and/or system behaviour. 15 In one advantageous refinement, it is in this case predictable that the predictive predetermination of at least one lower and/or one upper alarm limit and/or the respective limit value adaptation or alarm limit adaptation can be carried out in an automated form. 20 According to the invention, it has accordingly first of all been recognized that, for more efficient monitoring, alarm limits and the limit values relating to them must be selected such that they cover all the values which are permitted for an installation which is operating correctly 25 and without disturbances respectively, or for a supply system. At the same time, however, determined values, in particular measured, simulated or calculated values which should be considered to be critical cannot be ignored. It has also been found that statically defined constant alarm 30 limits and limit values conceal the risk of ranges for permissible values to be determined, in particular values to be measured, to be simulated or to be calculated, being selected to be too narrow. However, it has also been found that excessively non-specific definition of alarm limits 35 and/or limit values can result in critical states of an 2448954_1 (GHOtatters)25/10/2010 - 6 installation not being detected in good time and/or reliably. Variable and dynamic matching to the respective conditions of a system is achievable by varying or matching an alarm 5 limit and/or an alarm limit profile. The respective matching and the matched alarm limit resulting from this, or the alarm limit profile resulting from this, in this case takes account of the different profile of permissible determined values, and avoids false alarm signals. To this 10 extent, efficient monitoring and low-disturbance operation resulting from this in an installation thus protected by alarm are permanently feasible. The object mentioned initially has thus been achieved. Against this background, it is considered advantageous for 15 at least one lower and/or one upper alarm limit and/or an alarm limit profile based on this to be predetermined predictively, on the basis of and with assessment of historical and/or empirical and/or simulated and/or calculated process and/or time-series information items, 20 in particular process values as well as environmental and/or state conditions associated with them. A further refinement provides that at least one lower and/or one upper alarm limit is variable and/or adaptable as a function of at least one process-internal or process 25 external peripheral variable. A peripheral variable is not a value to be determined, in particular to be measured or which has been measured and which is compared with the respective limit values, but a variable which represents external or internal factors which influence the monitored 30 process. This allows the alarm limits to be dynamically matched to process-external or process-internal changes which influence the values to be measured. In one preferred refinement, the respective lower and/or upper limit value has a different magnitude in a 2448964_1 (GH1atLero)25/10/2010 predeterminable first time interval, in particular in the case of a value profile which is to be measured and/or to be detected and varies over time, than in a predeterminable second time interval which follows the 5 first. This ensures that the limit value or limit values can follow an already known or defined time profile of permissible values to be measured or to be detected. This makes it possible to take account of and/or to very largely compensate for, for example, fluctuations that are 10 dependent on the time of day, the day of the week or the time of year, as well. The respective lower and/or upper limit value profile or the respective lower and/or upper alarm limit is particularly preferably determined predictively in that 15 the respective characteristic variables and/or measurement variables as well as the values relating to them are determined predictively and/or as a function of time, as a function of a prediction of external environmental states or internal process states, and a lower and/or an upper 20 limit value is defined on the basis of the determined characteristic variables and/or measurement variables for the respective time and until the next prediction. A corresponding limit value profile can therefore be determined and/or defined over the various points in time, 25 and also within a predeterminable interval. This specific refinement makes it possible to use a profile, which is dependent on the environment and is already known, in particular already historically known, for the values to be measured as a basis for determining 30 the limit values and/or the respective alarm limits. A predictable and expected fall or rise in the values to be measured can in this way be taken into account and/or used and compensated for, without false alarm signals being triggered. 2448964_1 (GHMattero)25/10/2010 -8 In a further advantageous embodiment, the characteristic variables and/or measurement variables as well as the values relating to them are estimated for predictive determination of alarm limits, in particular by prediction 5 of external environmental states or internal process states, wherein the future profile of measurement, simulation or calculation variables and/or of values relating to them is determined using statistical methods, in particular signal processing, expert systems and/or 10 empirical values, and the respective alarm limits are defined on the basis of them. A time-based prediction of the variable which is actually to be monitored is accordingly produced and in this case may be any desired physical variable, a measured value, a 15 calculated value or a simulated value, and the alarm limits are determined, in particular calculated and defined, on the basis of the respective prediction, or the future upper and/or lower alarm limits and/or alarm limit profiles are determined. 20 In a further preferred refinement, the at least one dynamically adaptable lower and/or upper limit is determined and defined, and in particular predicted, as a function of predicted and/or predictively determined characteristic variables and/or measurement variables to 25 be detected and/or measured, as well as values relating to them. As a development, the characteristic variables and/or measurement variables as well as the values relating to them can be estimated for prediction, and/or simulated, 30 calculated or else estimated values can be used as predicted values. In this case, empirical and/or historical information and data, in particular as well as time-series information, can once again be used for simulation and/or estimation. 2448964_1 (GHIattera)25/10/2010 -9 Against this background, the characteristic variables and/or measurement variables to be measured and/or to be determined, as well as values relating to them, are defined predictively at the start of one or more time 5 intervals, and in particular are determined statistically. A model which is used for predictive limit value determination, and with the aid of which a future value is determined predictively, must in this case be sufficiently accurate to reliably map and/or simulate the future 10 profile of the actually measured values, and in particular to estimate or assess them in advance. This model can be described by mathematical-physical equations, data-based methods, such as fuzzy logic, neural networks, Kalman filters, support vector machines, etc., learning systems 15 or other methods. Historical data and/or expert knowledge or else further forecast data and forecast information, such as weather data, can be used as a basis for an estimate. Optimized upper and lower alarm bands can be defined by 20 the introduction of expert knowledge and/or signal processing methods and/or with the assistance of learning systems and/or with the assistance of system models, after a certain operating time or by objective analysis of historical and/or empirical data, with an alarm band 25 denoting the separation between a predicted value and a lower and/or upper dynamic alarm limit, an optimized separation between two points in time at which an estimation process is carried out, and further parameters are set. In this case, during a training period, for 30 example, the installation operator may if required dynamically adapt and/or adjust an alarm band determined in this way, on the basis of his experience and by commenting on alarms, such as true alarm, false-positive alarm, false-negative alarm. 2448964 1 (GH1attero)25/10/2010 - 10 The predictively determined, in particular also predicted, limit values and therefore the dynamic alarm limits can take considerably better account of external environmental states or internal process states, and can therefore also 5 reduce the number of alarm signals, and in particular false alarms, and in consequence can improve the alarm quality. The abovementioned procedure and the alarm limits and/or alarm limit profiles which are formed predictively take 10 account of the fact that, at any time, and also in the future, there is an environmentally dependent area with permissible values in the form of a band, in such a way that no alarms are triggered for measured values which are within the band. 15 in a further refinement, at least one dynamic lower limit value, which is variable, is compared with at least one further lower, static limit value, and/or at least one dynamic upper limit value, which is variable, is compared with a further upper, static limit value. The one upper, 20 static limit value in this case represents an absolute maximum value which must never be overshot in any circumstances, and also not by a dynamically adaptable limit value or an alarm limit relating to it. The lower, static limit value in this case represents an absolute 25 minimum value, which must never be undershot in any circumstances. This ensures that determined, variable limit values always include process-critical limits. If the static maximum value is overshot or the static minimum value is 30 undershot, an alarm signal is always produced. In order to prevent the number of alarm limits from being reduced at those points where the upper and lower dynamic alarm limit values respectively overshoot or undershoot the static alarm limit value, a minimum separation, which must 35 additionally be complied with, can also be defined between 2448964_1 (GHMattera)25/10/2010 - 11 the dynamic upper alarm limit and the static upper alarm limit, and between the dynamic lower alarm limit and the static lower alarm limit. Particularly preferably, the measured, simulated and 5 calculated values are compared with a respective first lower, variable/dynamic limit value and a second lower, static limit value and/or with a respective first upper, variable/dynamic limit value and a second upper, static limit value. This ensures that an alarm signal is produced 10 when the measured values overshoot a maximum value or undershoot a minimum value. By way of example, the determined values may in this case be pressures and/or flow rates and/or flow velocities and/or temperatures, in particular of liquids such as 15 water or oil or other chemical substances, or else of gases. This also allows the method described here to be used in the field of water supply companies or water supply systems, as well as the chemical industry, and in the field of oil and gas. Other application areas are also 20 feasible, for example for power distribution and/or power generation, or else in general for process automation. Furthermore, the object mentioned initially is achieved by a system for efficient monitoring of a process and/or of a technical installation, having a monitoring device which 25 determines characteristic variables and/or measurement variables for the respective process and/or for the respective technical installation and detects, in particular measures, simulates or calculates, values relating to them, and compares them with predictively 30 determined alarm limit values of at least one lower and/or one upper alarm limit, and the monitoring device generates an alarm if a lower alarm limit is undershot or if an upper alarm limit is overshot by a determined value, in particular by a measured, simulated or calculated value, 35 wherein at least one lower and/or at least one upper alarm 2448964_1 (GHMattero)25/10/2010 - 12 limit is variable on a state-basis and/or as a function of the environment and is dynamically adaptable, in particular according to the process behaviour and/or installation behaviour, by means of the monitoring device. 5 In one advantageous refinement, the respective alarm limit adaptation and/or limit value adaptation relating to it can be carried out or prompted in an automated manner by means of the monitoring device. Provision can also advantageously be made for the 10 determined and/or adapted alarm limit values and/or alarm limit profiles relating to them to be predeterminable and/or externally definable, for example by operating personnel. I(n a further refinement, at least one alarm limit 15 transmitter device is provided by means of which at least one lower and/or at least one upper alarm limit is predictively predeterminable. The alarm limit transmitter device is in this case advantageously integratable in the monitoring device. 20 In a further refinement, the monitoring device has at least one data processing device, in particular in the form of a microcontroller, a microcomputer, a PLC or an ASIC. Furthermore, advantageously, input and/or display means 25 can be provided for manual detection and influencing of limit values, and for indicating and/or displaying them. In addition, a further advantageous embodiment of the system comprises at least one interface for wire-free communication, in particular by means of Bluetooth and/or 30 WLAN, and/or cable-based communication, in particular by means of USB, Ethernet, RS232, SCSI, Profibus, Profinet and the like, having at least one control system, in 24489641 (GHatters)25/10/2010 - 13 particular a process control system, as well as one or more measurement points. In addition, a data memory can advantageously be provided in which historically and/or empirically determined time 5 series information as well as data are stored such that they can be called up. In one advantageous refinement, the abovementioned apparatus and in particular its monitoring device comprise means for carrying out at least one of the method versions 10 described above, as a result of which, in order additionally to avoid repetitions relating to the inventive step, reference is made to the statements relating to the method, as such. In one advantageous refinement, the abovementioned 15 apparatus is designed and prepared to carry out all the Steps individually or in combination, as described and/or claimed by the method. The method described here and the apparatus described here are in this case usable in particular for monitoring so 20 called "soft alarms" or "smart alarms" which, for example, are produced at the control system level in a comprehensive monitoring system, in particular in a control station. However, in principle, alarms can also be monitored at the controller level, that is to say by 25 appliances such as PLCs, RTUs and the like which are used at the monitoring and/or control level. The rest of the description of the invention will be based on a number of figures and exemplary embodiments. Preferred refinements and developments for the teaching 30 will also be explained, in a general form, in conjunction with the explanation of preferred exemplary embodiments of the invention with reference to the drawing. In the drawing: 2448964_1 (GHMatters)25/10/2010 - 14 Figure 1 shows a time-value graph in which the values plotted as a function of time undershoot a first, lower static limit value, Figure 2 shows a time-value graph in which a lower limit 5 value is reduced in a specific time interval, Figure 3 shows a time-value graph in which the limit values are calculated dynamically, and Figure 4 shows a time-value graph in which values to be measured are predicted by estimation at specific times. 10 The time-value graph shown in Figure 1 describes a method for monitoring a process, in which values are determined, in particular measured, simulated or calculated, and are compared with lower and upper alarm limit values which have been set, with an alarm signal being produced if a 15 determined value, in particular a measured, simulated or calculated value, undershoots a lower alarm limit value or overshoots an upper alarm limit value. Figure 1 shows a time-value graph according to the prior art, in which the values determined by way of example 20 undershoot a first lower static alarm limit value. In this time-value graph (time series), the alarm limit values, specifically a first and a second upper alarm limit value and a first and a second lower alarm limit value, are defined statically. The alarm limits are in this case 25 predetermined and defined on the basis of in each case one fixed or constant limit value, in which case manual adaptation is possible in principle, for example in the course of the engineering process. As shown, a determined value undershoots the first lower 30 alarm limit value at the time tO. This results in an alarm signal. This alarm signal may, however, in some circumstances be a false-alarm signal. 2448964_1 (GHMattero)25/10/2010 - 15 The graph in Figure 2 shows a method for monitoring a process in which values are determined, in particular measured, simulated or calculated, and are compared with lower and/or upper alarm limit values which are set, in 5 which case an alarm is generated and/or signalled when a determined value, in particular a measured value, undershoots a respective lower limit value or overshoots a respective upper limit value. The lower alarm limit value can be dynamically adapted and/or varied on a state basis 10 and/or as a function of the environment during the time dependent measurement, simulation and/or calculation in progress. Figure 2 shows a time-value graph (time series), in which the lower alarm limit value is decreased within a 15 predeterminable time interval. In this exemplary embodiment, the lower limit value is reduced in the time interval tO to t2, since the values determined and/or measured in this interval are permissible for a correct process profile. The lower limit value therefore has a 20 different alarm value in a first time interval than in a subsequent, second time interval. No false-alarm signal is produced, or a false-alarm signal such as this is avoided, by matching the alarm limits to the prevailing circumstances at the time tl, thus making it possible to 25 sensibly reduce the number of alarms. The limit values and/or the alarm limits are periodically and/or cyclically redetermined and accordingly adapted or varied as a function of time and event, as a function of and/or using process-external or process-internal 30 peripheral variables, simulations or calculations. These newly determined, in particular calculated, limit values are entered and/or applied to the previous alarm limits in an automated form. This allows alarm signals, in particular "soft alarms" or "smart alarms", to be produced 35 or generated as a function of dynamically varying limit 2448964_1 (GHO4ters)25/10/2010 - 16 values and/or alarm limits, and to be configured efficiently. In a further refinement, the determined, in particular measured, simulated or calculated, values and the 5 respective limit values in Figure 2 can also represent, for example, pressures, flows or temperatures, such as those monitored in the case of a water or gas supply company, the chemical industry or in power distribution and/or power generation. 10 A lower limit value for the pressures of a gas or of a liquid could, for example, be 4 bar, and an upper limit value could be 8 bar. An alarm or corresponding alarm signal is generated when a measured value of a pressure is less than 4 bar or is more than 8 bar. 15 However, if the demand is lower, a smaller lower limit value could be sufficient to ensure reliable supply, in particular gas or water supply. By way of example, this is the case at night, when fewer consumers require gas or water. A supply company can adapt the pressure, for 20 example in order to reduce the leakage rate and to reduce the load on the respective line system. Nevertheless, it should still be possible to produce reliable alarm signals, and to monitor the state of the system, at all times, thus allowing the respective installation and/or 25 the system to be operated within the permissible operating limits, which are governed by the design of the process and/or of the respective installation and/or of the supply system. In addition, predictive determination and dynamic 30 adaptation of the alarm limits makes it possible, for example, to efficiently take account of and assess varying demand levels, flows, pressures, powers, currents and the like on the basis of holiday times, vacation times and journey times, public holidays, summer time and winter 35 time and the like. 2448964_1 (GHMattero)25/10/2010 - 17 In the specific example described here, the lower limit value can automatically be reduced from 4 bar to 3 bar during the period from tO to t1. Therefore, an alarm signal is triggered only if the measured pressure falls 5 below 3 bar. This system-related, state-based and event-dependent adaptation of the limit value has the advantage that fewer alarms are produced, and the alarms better reflect a critical state of the system. Dynamic variation of the 10 limit values as a function of external or internal factors has the positive effect that false alarms which result from permissible fluctuations in the measured values or values to be measured now no longer occur, or occur at least to a reduced extent. 15 Figure 3 shows a time-value graph which shows the profile of a forecast, estimated, calculated or simulated measurement variable or variable to be monitored, as well as the alarm limits relating to it. The respective alarm limit values and/or alarm limits can likewise be forecast 20 as a function of forecast external and/or internal factors. In this case, defined first upper and first lower limit values or alarm limits are associated with a predeterminable time interval. The respective lower and the respective upper alarm limit value are determined and 25 predicted as a function of predicted values to be measured. Specifically, Figure 3 shows how the limit values vary as a function of predicted values to be determined (measurement variable or system state). The limit values 30 have a profile at equidistant intervals parallel to the respective predicted value to be measured. An interval between the predicted value to be measured and the respectively associated calculated limit value is referred to as the alarm separation. 2448964_1 (GHMattero)25/10/2010 - 18 The magnitude of the alarm separation between the first upper limit value and the predicted value need not in this case be identical to the magnitude of the alarm separation between the lower first limit value and the predicted 5 value, and in addition the alarm separations need not be the same over the entire time period. A second upper and a second lower static limit value ensure that a critical system state can never be overshot during a process. If the predicted respective first upper 10 limit value is above the respective second upper limit value, as is the case in the time interval between tl and t2 in Figure 3, then the second upper limit value is considered to be a maximum value which must not be exceeded. A corresponding situation applies to the second 15 Lower limit value in the time interval t3 to t4. A minimum separation can be defined, which is the minimum which must be maintained between the dynamic alarm limit value and the static limit value. For this purpose, the first lower limit values, which are 20 variable or dynamically adaptable, are compared with a second lower, static limit value. The first upper limit values, which are variable and/or dynamically adaptable, are compared with a second upper, static limit value. :Depending on the configuration of a system, the second 25 upper limit value or the second lower limit value, or even both limit values, may be cancelled, or further dynamic and/or static alarm limit values and/or alarm limits may be added and considered. It is also feasible to define a maximum and minimum, which 30 must never be overshot or undershot, instead of a second upper limit value and instead of a second lower limit value or a predeterminable alarm limit for the first upper and first lower limit values. 2448964_1 (CHMatters)25/10/2010 - 19 The method as shown in Figure 3 is used in particular in the field of supply facilities such as water or gas supply, or else in the chemical industry, but can also be used in and applied to virtually all other technical 5 fields, such as process automation, power generation, power distribution, power supply and in the field of power stations. Particularly in the field of water supply, comparatively major fluctuations can occur in the hourly and/or daily water consumption. A future water consumption 10 can be calculated in advance, and therefore predicted, with the aid of historical data, empirical values and suitable signal processing methods. These advance calculations and predictions are used by the respective supply company in particular to optimize the 15 operation of, for example, pumping stations or substations and the provision of the required water, gas or respectively required energy at the respectively required pressure or the respective voltage. The limit values for the required power (current and voltage) , flow rates and 20 pressures in the respective supply lines can now be calculated in advance, and therefore dynamically adjusted in a worthwhile manner, as a function of the predicted consumption. A future water consumption to be expected, for example, 25 can in particular be determined on an hourly basis. The future limit values for flow rates and pressures which will occur in a future time interval are also calculated and predicted with each new prediction relating to the water consumption. 30 Figure 4 shows a time-value graph in which values to be measured at specific times (measurement variables or system state) are predicted by estimation and/or based on historical and/or empirical information. Values to be measured or measurement variables are frequently not 35 estimated continuously, but only at specific times. This 2448964_1 (Gtattera)25/10/2010 - 20 may also be the case, or may be applicable, for simulated and calculated values. The respective first lower and the respective first upper limit value profile are calculated and predicted as a 5 function of estimated values to be measured. The values to be measured are estimated at the start of one or more time intervals. In Figure 4, the value actually to be measured, to be simulated or to be calculated is estimated at the times 10 t0, t1, t3 and t5. These times are fixed. If the actually measured value differs excessively from the estimated or predicted value, this is identified, and a new estimate can be initiated. This estimate takes account of historical information or signal models, as a result of 15 which a newly developing fault in the process, which is leadingg to a change in the measured, simulated and/or calculated variable will influence the estimated limit value profile only to a limited extent, and therefore an alarm is triggered. The consideration of historical 20 :information or signal models prevents the dynamic limit value profiles from automatically being matched to a fault in the process/system, in which case an alarm would be triggered only very late. In addition, where necessary, Lhe measured, simulated and/or calculated signals are 25 reprocessed, whereby undesirable signal characteristics, such as spurious values, are identified and corrected. Furthermore, this preprocessing of signals also makes it possible to identify trends in the signals, and thus to identify slowly developing faults in the process and 30 prevent the dynamic alarm limits from drifting away slowly. A minimum time period At can be defined between two estimates, in which the estimated dynamic limit values must not be recalculated, in order to prevent continual recalculation in the event of a fault. Instead of this, in 35 this case, an appropriate alarm can be triggered, since this indicates a fault in the process. 2448964_1 (GKMtter)25/10/2010 - 21 If the measured, simulated and/or calculated value overshoots the upper limit value after a new estimate, or if it undershoots the lower limit value, an alarm is triggered since there is then a fault in the process 5 and/or operation. Estimates must be trustworthy and must assist operation of technical installations, that is to say a discrepancy from the estimate of a value to be determined results in an alarm. The forecast or predicted limit values in the time 10 interval tO to tl are based on the estimate of a predicted value, which is actually to be measured or is to be determined in some other way, at the time to. _n reality, the actually measured and/or determined value follows the estimated or predicted value only within a 15 certain tolerance range. The alarm separations between estimated values and limit values are selected such that permissible discrepancies between estimated values and actually measured values do not trigger false-alarm signals, and desired alarm signals are produced reliably. 20 For example, the predicted values can be based on (signal) models for simulation, which represent the ideal state, that is to say that an "undesired" or "undesirable" change in the process may form the basis for a discrepancy between the actual value and the predicted value, in which 25 case a comparatively high simulation model quality must be ensured. For this purpose, the lower and/or upper alarm limit values are selected and calculated as a function of predicted values to be measured, such that the difference 30 between a predicted value to be measured and an upper alarm limit value is greater than zero, and the difference between a predicted value to be measured and a lower alarm limit value is less than zero. The measured values are compared with a respective first 35 lower, variable limit value and a second lower, static 2448964_1 (GWD4atters)25/10/2010 - 22 limit value, and/or with a respective first upper, variable limit value and a second upper, static limit value. It is possible to add further dynamic and/or static alarm limit values and/or alarm limits to the system, and 5 to consider them for alarm generation. With respect to further advantageous refinements and developments of the teaching according to the invention, reference is made on the one hand to the general part of the description, and on the other hand to the patent 10 claims. The present invention also covers any combinations of preferred embodiments or developments, provided that they are not mutually exclusive. In the claims which follow and in the preceding 15 description of the invention, except where the context requires otherwise due to express language or necessary implication, the word "comprise" or variations such as "comprises" or "comprising" is used in an inclusive sense, i.e. to specify the presence of the stated features but 20 not to preclude the presence or addition of further features in various embodiments of the invention. It is to be understood that, if any prior art publication is referred to herein, such reference does not constitute an admission that the publication forms a part of the 25 common general knowledge in the art, in Australia or any other country. 2448964_1 GIOMaters)25/10/2010

Claims

1. Method for efficient monitoring of a process and/or of a technical installation and/or of a supply system, wherein at least one lower and/or at least one upper alarm 5 limit is determined predictively, and/or characteristic and/or measurement variables as well as values relating to them for the respective process and/or for the respective technical installation are determined and are compared with predictively determined alarm limit values of at 10 least one lower and/or at least one upper alarm limit, and wherein if a lower alarm limit is undershot or an upper alarm limit is overshot by a determined variable and/or a value determined in a manner relating to it, in particular a measured, simulated or calculated value, then an alarm 15 signal is generated, and wherein at least one lower and/or at least one upper alarm limit is variable on a state basis and/or as a function of the environment and is dynamically adaptable, in particular according to the process behaviour and/or installation behaviour. 20 2. Method according to Claim 1, characterized in that the predictive predetermination of at least one lower and/or one upper alarm limit and/or the respective limit value adaptation and/or alarm limit adaptation can be carried out in an automated manner. 25 3. Method according to Claim 1 or 2, characterized in that at least one lower and/or one upper alarm limit is variable and/or adaptable as a function of at least one process-internal or process-external peripheral variable.

4. Method according to one of Claims 1 or 2, 30 characterized in that at least one lower and/or upper alarm limit value is variable as a function of at least one measured and/or simulated and/or calculated value of a characteristic variable or measurement variable. 2446964_1 (G60fttere)25/10/2010 - 24 5. Method according to one of the preceding claims, characterized in that the determination of the characteristic variables and/or measurement variables, as well as their values, can be carried out by means of 5 measurement, model-based simulation and/or calculation.

6. Method according to one of the preceding claims, characterized in that the determination of the characteristic variables and/or measurement variables, as well as their values, can be initiated and/or carried out 10 on an event-controlled basis, a time-controlled basis and/or by manual instruction.

7. Method according to one of Claims 1 to 5, characterized in that at least one lower and/or upper alarm limit value or an alarm limit is determined 15 predictively in that the respective characteristic variables and/or measurement variables as well as the values relating to them are determined predictively and/or as a function of time, as a function of a prediction of external environmental states or internal process states, 20 and a lower and/or an upper alarm limit value is defined on the basis of the determined characteristic variables and/or measurement variables for the respective time and until the next prediction and/or a predeterminable time and/or within a predeterminable time interval. 25 3. Method according to one of Claims 1 to 7, characterized in that at least one lower and/or one upper alarm limit value is determined, in particular predicted, as a function of predicted and/or predictively determined characteristic variables and/or measurement variables as 30 well as values relating to them.

9. Method according to one of the preceding claims, characterized in that the characteristic variables and/or measurement variables as well as the values relating to them are estimated for predictive determination of alarm 35 limits, in particular by prediction of external 2448964_1 (GHNaters)25/10/2010 - 25 environmental states or internal process states, wherein the future profile of measurement, simulation or calculation variables and/or of values relating to them is estimated using statistical methods, in particular signal 5 processing, expert systems, historical information and/or empirical values, and the respective alarm limits are defined on the basis of them.

10. Method according to one of the preceding claims, characterized in that the characteristic variables and/or 10 measurement variables to be determined as well as values relating to them are determined predictively at the start of one or more time intervals.

11. Method according to one of the preceding claims, characterized in that the lower and/or upper alarm limit 15 values are selected and determined, in particular calculated, such that the magnitude of the difference between an alarm limit value and a predictively determined future value is not equal to zero.

12. Method according to one of the preceding claims, 20 characterized in that first lower alarm limit values, which are changed, are compared with at least one second lower, static alarm limit value, and/or in that first upper alarm limit values, which are changed, are compared with at least one second upper, static limit value, and/or 25 are dynamically adapted such that the difference between the respective dynamic and static alarm limit values, in particular the respective alarm limits, does not undershoot or does not overshoot a predeterminable minimum separation. 30 13. Method according to one of the preceding claims, characterized in that the determined values, in particular measured and/or simulated and/or calculated values, are compared with at least one first lower, variable alarm limit value and with at least one second lower, static 35 alarm limit value and/or with a first upper, variable 2448964_1 (GHMatters)25/10/2010 - 26 alarm limit value and with at least one second upper, static alarm limit value, and if at least one upper alarm limit is overshot and/or at least one lower alarm limit is undershot, an alarm is generated, in particular an alarm 5 signal is triggered and/or an alarm message is output.

14. Method according to one of the preceding claims, characterized in that the characteristic variables and/or measurement variables determined and/or to be determined as well as values relating to them comprise pressures 10 and/or flow rates and/or temperatures and/or at least one undefined other physical variable.

15. System for efficient monitoring of a process and/or of a technical installation, and/or of a supply system, having a monitoring device which determines characteristic 15 variables and/or measurement variables for the respective process and/or for the respective technical installation and detects, in particular measures, simulates or calculates, values relating to them, and compares them with predictively determined alarm limit values of at 20 least one lower and/or one upper alarm limit, and the monitoring device generates an alarm if a lower alarm limit is undershot or if an upper alarm limit is overshot by a determined value, in particular by a measured, simulated or calculated value, wherein at least one lower 25 and/or at least one upper alarm limit is variable on a state basis and/or as a function of the environment and is dynamically adaptable, in particular according to the process behaviour and/or installation behaviour, by means of the monitoring device. 30 16. System according to Claim 15, characterized in that the monitoring device signals alarm limit overshoots and/or alarm limit undershoots.

17. Apparatus according to Claim 15 or 16, characterized in that means are provided for carrying out the method 35 according to one of Claims 1 to 14. 2448964_1 (GHMattero)25/10/2010