HUP0200360A2 - Heating installation and method for operating the same - Google Patents

Abstract

The subject of the invention is a heating device, which has a storage tank (6), which has an outlet and inlet pipe for the heated liquid, a supply arrangement (9, 10) for the heat transfer liquid in a supply pipe equipped with a valve (11), and a temperature sensor (13) which measures the temperature of the heated liquid. , and with a control circuit (18), which operates the valve as a function of the deviation of the temperature from the prescribed value, as well as a procedure for operating this equipment. In doing so, we want to achieve a quick reaction of the equipment with a low mechanical load on the moving parts. For this, the control circuit (18) is equipped with a limit frequency sensor (19), which detects temperature fluctuations (Tist) and reduces the loop gain (V) of the control circuit (18) at too high a frequency, and if the frequency is too low, it increases. HE

Description

73.071/ΚΟΤ

PUBLICATION COPY

Heating device and method for operating the same

The invention relates to a heating device, which has a storage tank, which has an outlet and inlet pipe for heated fluid, a supply arrangement for heat transfer fluid guided in a supply pipe equipped with a valve, a temperature sensor, which measures the temperature of the heated fluid, and a control circuit, which operates the valve depending on the deviation of the temperature from the set value. Furthermore, the invention includes a method for operating a heating device, in which the temperature of the heated fluid is measured, and the addition of heat transfer fluid is controlled by means of a control circuit valve depending on the deviation of this temperature from the set value.

The invention is presented below on the basis of a device for providing hot water. However, the invention can also be used in other heating systems in which radiators or underfloor heating are to be supplied with the liquid. A known example of a heating system is the patent DE 41 42 547 A1, which also serves both purposes. The present invention is based on this. The control of a circulation pump in a heating system is known from the patent DE 24 52 515 A1. Here too, a similar principle was used.

The heat transfer fluid does not necessarily have to be water, it can even be a gaseous substance, such as heated air that supplies a room or room system.

The heated fluid receives its heat from the heat transfer fluid. The heat transfer fluid can flow through a heat exchanger, on the other side of which is the heated fluid. It is also possible for the heat transfer fluid to be heated directly by a heat source, such as a burner, and then mixed with the heated fluid.

However, a common feature in all cases is that the heat transfer fluid is fed in by a valve. If the temperature of the heated fluid drops, heat must be supplied by opening the valve that controls the flow of the heat transfer fluid. If the temperature then rises above the specified value, the valve must be closed again. In most cases, the valve is located in the heat transfer fluid supply pipe. However, this is not a requirement as long as the valve is able to control the amount of heat delivered by the heat transfer fluid. The valve can also be located in the discharge pipe.

The temperature of the heated liquid does not necessarily have to be measured in the storage tank. It is also possible to measure this temperature in the pipe leading from the storage tank to the actual heating circuit. In extreme cases, the storage tank itself can be relatively small, or even form the heating system itself.

In systems of this type, we are faced with the following contradictory phenomena. On the one hand, we want to keep the temperature of the heated fluid at the required value quite well. So, if there is a heat demand, for example, when we draw hot water from a tap, the heat transfer fluid must replace the removed heat as quickly as possible in order to heat the cold water that is being refilled. On the other hand, it has been noticed that with fast control circuits, these circuits tend to oscillate. In a hot water production system, this has really noticeable consequences. The water temperature fluctuates. Temperature fluctuations are less critical in a heating system that only supplies radiators or even just underfloor heating. However, in this case, too, the oscillation leads to a significant load on the valve or the stepper motor that drives the valve.

The invention is based on the task of achieving a rapid response of the heating device under low mechanical load.

This task is solved in the heating system described in the introduction by having a limit frequency sensor in the controller, which detects temperature fluctuations and reduces the loop gain of the controller if the frequency is too high, and increases it if the frequency is too low.

With this design, we obtain a self-adaptive loop gain of the control circuit, so that on the one hand we avoid oscillations whose frequency exceeds an acceptable level, and on the other hand we always get a relatively fast response of the heating system in case of heat demand. This is not about avoiding temperature oscillations at all. It is simply about keeping the loads that affect the mechanical control elements, such as the valve or the drive, low so that the service life is not reduced too much. The loop gain of the control circuit can be adjusted relatively easily by changing the static gain of the controller. To a first approximation, it is inversely proportional to the static gain of the controller.

It is preferable if the cut-off frequency detector has a threshold element and the frequency is detected by the output of the threshold element.

In other words, it only detects oscillations during frequency measurement whose amplitude exceeds the threshold value. This creates a band around the specified value in which any oscillations can occur without affecting the frequency measured by the cut-off frequency sensor. The cut-off frequency sensor therefore only measures oscillations that fall outside this band.

It is advantageous if the limit frequency sensor detects temperature fluctuations indirectly from the valve's actuation signal or from the valve movement. In principle, the controller only moves the valve when necessary, in other words, when the temperature deviates from the set value by a predetermined amount. If such a deviation has occurred, the difference is already available. It then appears as valve movement, or, which is easier to perceive, as a signal that triggers valve movement. In other words, we use information that is already present in the control circuit.

In the method described in the introduction, the task is solved by determining the frequency of temperature fluctuations and reducing the loop gain during regulation at too high a frequency, or increasing it at a sufficiently low frequency.

As explained earlier, this way we get a control circuit that works with the highest possible loop gain without starting into impermissible oscillations. This gives us a very fast response and at the same time protects especially the mechanical components, such as the drive and valves. The loop gain is adjusted adaptively, which means that it adapts exactly to the current situation.

So it can vary from device to device, and within a device from hour to hour.

It is preferable to measure only the frequency of deviations that exceed a predetermined difference from the set value. This creates a tolerance band around the set value, within which oscillations of any frequency are permitted. These oscillations do not place excessive strain on the actuators, because the adjustment movements occurring here are very small. For any user drawing hot water, oscillations within the band are hardly noticeable and are therefore acceptable.

It is advantageous to determine the frequency indirectly from the valve actuation signal and/or the valve movement. The actuation signals and the resulting valve movements are direct consequences of the temperature deviation from the setpoint. This means that we have information about the deviation and it is expressed in relatively easy-to-measure signals. These can then be evaluated with relatively little effort.

It is useful to measure the number of valve reversals over a predetermined time period and reduce the loop gain if the number exceeds a maximum value. Frequency measurement can be limited to a simple count, where the count must occur over a predetermined time period. If this predetermined time period is set to, for example, 5 minutes, then a specified number of valve reversals, for example 3-10, can be allowed without detecting instability. If more reversals than planned occur, the system is considered unstable and the loop gain is reduced.

It is particularly advantageous to interrupt the counting and start the time range again whenever an overshoot occurs. This allows us to reach a stable state even faster. The more unstable the system, the higher the frequency, and therefore the more often the valve changes direction. If we correct when the condition is met, we do not have to wait for the entire time range to be corrected. This reduces the load on the mechanical components and allows us to reach a stable state much faster.

In the case where the frequency is quite low, it is advantageous to increase the loop gain and change the setpoint. So, not only do we increase the loop gain, but we also change the setpoint to determine whether the system starts to oscillate, i.e. the control time. In a non-oscillating system, even increasing the loop gain does not automatically lead to oscillations, so we cannot be sure that the loop gain is adequate. However, by changing the setpoint, we produce a jump that provides the desired information.

It is advantageous to use the value of the loop gain before the increase in the case where the loop gain has been increased and the frequency has proven to be too high. This allows us to reliably probe the “limit”. For the same load, we have information about which loop gain is still stable at the control time, and we also have information about which loop gain is no longer stable at the control time for the next increase. We can then return to the previous loop gain without having to perform another iteration.

In a particularly preferred embodiment, the loop gain is given as a function of the load on the system. This is a further possibility of protecting the control system and the moving elements therein from being overloaded by too frequent movements. If the demand on the system is low, for example, only a small amount of hot water is drawn off, a low control gain can be used. A fast reaction is not required. The same applies, for example, to night-time regulation, when the radiator valves of a heating system are regulated down, so that only a small amount of heat is “used up” or “discharged”. However, if there is a demand, for example, hot water is drawn off or the radiator valves are opened, then a fast reaction of the system is required. In this case, a higher loop gain can be switched to. In this case, when the demand is used as an additional criterion, a part of the iterative process can be skipped by several steps.

Preferably, the load on the system is measured by the temperature of the heated fluid. This method is sufficiently fast and does not require additional components. If the system is loaded, for example by drawing off hot water, the temperature in the storage tank drops relatively quickly by supplying a sufficient amount of cold water. Accordingly, the loop gain can be increased relatively quickly without the risk of oscillations developing immediately. If oscillations develop after a certain time, we can assume that the load has now ended and we can jump back to the “idle value” of the loop gain.

The invention will be described in more detail below with the aid of a preferred embodiment in accordance with the drawings. Based on this:

Figure 1: Schematic representation of a heating system for producing hot water,

Figure 2: Schematic representation of a control device,

Figure 3: different curves to illustrate the reduction of loop gain,

Figure 4: appropriate curves to represent the increase in loop gain,

Figure 5: Schematic diagram to explain system protection function.

Figure 1 schematically shows a heating system 1 which produces domestic hot water which can be drawn off at taps 2 or other outlets. The taps 2 are connected to a circuit 3 which consists of a supply pipe 4 and a return pipe 5 which are connected to a storage tank 6, for example a boiler. A circulation pump 7 is arranged in the circuit 3 which serves to make hot water available at the taps 2 with a negligible delay.

The storage tank 6 is designed as a heat exchanger, on the primary side 8 of which there is a supply pipe 9 and a discharge pipe 10 for a heat transfer fluid, or more generally a heat transfer fluid. The heat transfer fluid can be water, which is already provided by a heating boiler. However, it can also be a liquid that is used for heat transfer in district heating. The specific implementation of the heating of the heat transfer fluid does not play a significant role.

The valve 11 is located in the supply pipe 9 and can be opened and closed by means of a motor 12. The motor 12 is designed, for example, as a stepper motor, so that different opening states of the valve 11 can be set.

A temperature sensor 13 is arranged on the outlet pipe 4, which measures the temperature of the water in the outlet pipe 4. The temperature sensor 13 is connected to a control device 14, which in turn controls the motor 12. The control device 14 has an input 15, with which a setpoint value for the temperature in the storage tank 6 is given. This setpoint value is also called the “setpoint”.

If hot water is now drawn from a tap 2, cold water is simultaneously filled into the storage tank 6 through the supply pipe 16. The non-return valve 17 prevents water from flowing back from the circuit 3 into the supply pipe 16. The cold water naturally reduces the temperature of the hot water already in the storage tank 6. This temperature drop is detected by the temperature sensor 13. Based on this measurement, the control device 14 operates the motor 12, which opens the valve 11. The components mentioned therefore together form the control circuit 18. The control device 14 forms the actual "controller", which has a static gain Xp. The reciprocal of this static gain Xp is called the loop gain V.

Figure 2 schematically illustrates the more detailed structure of the control device 14. The inputs and outputs of the control device 14 are provided with the reference symbols to which the control device 14 is connected in Figure 1.

The control device 14 first of all has a differential amplifier 23, to which the setpoint value is fed from the input 15 and the actual temperature value from the temperature sensor 13. Depending on the difference between these two values, a corresponding control signal is provided for the motor 12. However, this 23

The static gain of the differential amplifier can be varied. A cut-off frequency sensor 19 is used for the variation. The cut-off frequency sensor 19 receives, first of all, the same signals as those received by the motor 12. It also receives the actual temperature and the nominal temperature. These signals or values are fed to a processing unit 20, which generates a pulse when certain conditions are met in a manner described later, and has, among other things, a threshold value element. The pulses are fed to a counter 21. The counter 21 is connected to a time base 22, which indicates to the counter 21 the beginning and end of a predetermined time interval. The output of the counter 21 is connected to the reset input of the time base 22. Furthermore, the output of the counter 21 is connected to the differential amplifier 23, more precisely to an input with which its gain, i.e. its static gain, can be adjusted.

The operation of the 18 regulators is described based on Figure 3.

Fig. 3a shows the T _ist curve, which is the temperature profile in the outlet pipe 4. The setpoint Tset, i.e. the nominal temperature, is indicated in dashes. Furthermore, the neutral zone Nz is indicated on both sides of the setpoint Tset. It can be seen that the temperature Tj _St initially fluctuates very strongly. The differential amplifier 23 generates pulses at specified intervals to operate the motor 12, which are shown in Fig. 3b. As long as the actual temperature Ti _St is lower than the nominal temperature Tset, the motor is moved in one direction (on+). If the situation is reversed, the motor is moved in the other direction (on-). This representation is of course only an example. Other methods of adjusting the valve 11 are of course also possible.

- 11 With the available heating system, we assume that repeated adjustment of valve 11 is not critical, as long as it is done in the same direction. We just want to avoid that the direction of movement of motor 12 and valve 11 changes too often to a greater extent.

In a manner not shown in more detail, the course of the adjustment direction is derived from the individual pulses, which are shown in Figure 3b. Figure 3c. Figure 3d. shows the output value of the counter 21, which value increases by one with each change of direction. The time element 22 gives a predetermined time range, which is indicated in Figure 3a. The time periods Zl, Z2, Z3 are, first of all, theoretically of the same length.

If it is now found during the time period Zl that the counter 21 has exceeded a predetermined value, first of all the static gain Xp of the differential amplifier 23 is increased, thereby reducing the loop gain V. At the same time the counter 21 is reset and the time base 22 is reset. Accordingly, the time period Z2 begins before the first time period Zl has completely run out. Here we use the fact that we do not even want to know how large the error actually is. It is enough to know that an error is present to initiate a correction.

Figure 3e now shows that the loop gain V is always reduced when the counter 21 reaches a predetermined count value, in this case 3, within a time interval Z. It can be observed that two corrections are required before the loop gain V becomes so small that the actual temperature Tist still fluctuates, but the amplitude of this fluctuation is mostly within the neutral zone.

In this case, only 2 changes of direction occurred during the time period Z3. Otherwise, the actual temperature Tist is

Tset - 0.5 Nz < Tist < 0.5 Nz + Tset requirement met.

The loop gain V is equal to the reciprocal of the static gain Xp of the differential amplifier 23. Now the following algorithm can be run. First of all, we define Xp:

Xp = (1+e) x Xp where é > 0 and is constant. Preferably é is also less than 1. After increasing the static gain Xp, which corresponds to reducing the loop gain V, the cut-off frequency sensor 19 is activated. If the cut-off frequency sensor detects instability, for example 3 or more direction changes in the time period Zl, Z2,...Zn, then the static gain Xp is increased once more, as described above. In the case where the number of direction changes does not reach the critical value, we assume that a stable state has been reached and maintain this static gain.

Basically, this procedure can be divided into two phases. In the first phase, we determine whether there are “critical oscillations” and at the end of the first phase, we change the loop gain accordingly. In the second phase, stability monitoring follows.

If no boundary oscillations are detected in the first phase, the loop gain is increased until an unstable counting period is observed. At this point, the adjustment process is interrupted and the gain of the previous stable counting period is selected. Figure 4 shows the procedure for increasing the loop gain. First, the static gain is reduced, for example, by

Xp = (l-h) x Xp, where h is a constant, greater than zero, and preferably less than 1. In the case of h, it can be the same value as é, but usually it will be a different value.

After this, we change the nominal value of the temperature Tset

Tset = Tset + Asp.

For this purpose, the control device has a setpoint transmitter 24 and a switchable sign-changing amplifier 25, which is connected to a summing point 26. The sign-changing amplifier is then switched, i.e. Asp = - Asp.

Changing the nominal value Tset results in a jump, which we want to induce oscillation. Without such oscillation, it would not be possible to detect instability even by changing the loop gain V. After changing the nominal value Tset, we wait for a delay time At. Then we start the cut-off frequency sensor 19. If and as long as the cut-off frequency sensor detects stable behavior of the controller 18, we continue this procedure, i.e. we increase the loop gain V.

However, at some point the loop gain V becomes so large that the controller 18 starts to oscillate. This is the case in the example of Figure 4 at time t3. In this case the nominal value Tset is reset to its original value and the static gain Xp is

Xp= l/(l-h) Xp is set to the value, thus completing the adjustment process.

The loop gain V is therefore reset to the value that allowed the last stable state.

If a stable state is reached in the adjustment process, as shown in Figure 3 or 4, no further adjustment occurs at first, the adjustment is complete. A further adjustment process only begins when another instability is detected, for example due to a load change that leads to the valve swinging.

Figure 5 shows another option for changing the loop gain V. This provides a system protection function that also avoids oscillations at low loads, close to idle.

In this case, we can distinguish between a low and a high loop gain, which can be switched between. The purpose of this protection function is to stabilize and optimize the heating system 1, depending on the load on the system.

The high loop gain V, denoted as V1 in Figure 5, is used for normal loading of the heater 1. This loop gain VI could be found, for example, during the automatic tuning process outlined above. Means not specifically described can be used to store this gain factor.

The low V2 loop gain is used in idle mode.

The cut-off frequency sensor 19 is used to determine the end of consumption. In this case, a high VI loop gain leads to an oscillation with too high amplitude and frequency. Therefore, if such an oscillation is detected after a preliminary increase in the loop gain, the gain is switched from the high VI value to the lower V2 value, after the system stabilizes again.

To determine the transition from idle to consumption, we use the temperature drop at the transition. This is, for example, the case shown at time t2 in Figure 5. Here, the draw-off takes place between time instants t2 and t3. We determine a load, in this case water consumption, if the actual temperature value falls below the nominal value Tset by the value Dz. In this case,

The loop gain is increased to the value VI. The controller thus obtains the speed required for normal consumption. At time t3, consumption ends. Since the loop gain is now too high, an oscillation occurs in the Z2 time domain. This is detected by the cut-off frequency sensor, and at time t4, the loop gain is reset to the value V2. The V2 loop gain can be found by increasing the loop gain in an appropriate iteration.

Claims (13)

. - ¹⁶ - λ Patent claims 1. Heating device with a storage tank, which has an inlet and outlet for heated liquid, a supply arrangement for heat transfer liquid guided in a supply pipe equipped with a valve, a temperature sensor, which measures the temperature of the heated liquid, and a control circuit, which operates the valve depending on the deviation of the temperature from the set value, characterized in that the control circuit (18) has a limit frequency sensor (19), which determines the temperature (T _is t) fluctuations and reduces the loop gain (V) of the control circuit (18) at too high a frequency and increases it in case of too low a frequency. 2. The device according to claim 1, characterized in that the limit frequency sensor (19) has a threshold element (20) and determines the frequency by taking into account the output of the threshold element (20). 3. Device according to claims 1 or 2, characterized in that the limit frequency sensor (19) determines the temperature (Ti _St ) fluctuations indirectly from the intervention signals or from the movement of the valve (11). 4. The device according to any one of claims 1 to 3, characterized in that the cut-off frequency sensor (19) has a direction change counter (21), a comparator, and a time base (22). 5. Method for operating a heating device, in which the temperature of a heated fluid is measured and, depending on the deviation of this temperature from the prescribed value, the addition of heat-transporting fluid is controlled by means of a control circuit valve, characterized in that the frequency of temperature fluctuations is measured and the loop gain is reduced if the frequency is too high or increased if the frequency is low enough. 6. The method according to claim 5, characterized in that only the frequency of deviations exceeding a predetermined difference from the prescribed value is measured. 7. The method according to claim 5 or 6, characterized in that the frequency is measured indirectly from the movement of the valve and/or its control signal. 8. The method of claim 7, characterized in that the amount of change in the direction of valve movement is measured over a predetermined time period and the loop gain is reduced if the value exceeds a maximum. 9. The method according to claim 8, characterized in that the counting is interrupted in the event of each exceedance and the time range is restarted. 10. The method according to any one of claims 5 to 9, characterized in that if the frequency is low enough, the loop gain is increased and the setpoint is changed. 11. The method according to any one of claims 5 to 10, characterized in that if the frequency is too high after increasing the loop gain, the value of the loop gain before increasing is reused. 12. The method according to any one of claims 5 to 11, characterized in that the loop gain is given as a function of the load of the device. 13. The method according to claim 12, characterized in that the load of the device is determined using the temperature of the heated liquid. The authorized person Dr. Zoltán Köteles, patent attorney Member of M S.B.G. & K. Patent Attorneys Office H-1062 Budapest, Andrássy út 113. Tel: 461-1000 Fax: 461-1099