HEATING SYSTEM WITH AUTOMATIC ADAPTIVE HYDRAULIC BALANCE

Description

[0001] The invention relates to a heating system with an automatic, adaptive, hydraulic balancing as well as a method for carrying out an automatic adaptive hydraulic balancing in a heating system and a method for training an artificial neural network.

[0002] Conventional heating systems used to regulate room temperature must undergo mechanical hydraulic balancing before commissioning. This involves adjusting the heating valves to ensure that all components of the heating system are coordinated with each other, thus ensuring even and satisfactory heating of the rooms.

[0003] Conventional hydraulic balancing requires the mathematical determination of target flow rates for each heating valve in the heating system during design and aims to prevent these values ​​from being exceeded during actual operation. For this purpose, defined hydraulic resistances are integrated into the hydraulic system, such as static throttle elements, whose setting values ​​are determined either from calculated target flow rates or from flow rates / differential pressures measured under defined load conditions, or dynamically controlled limiting devices, which are adjusted to the respective pressure or flow target values.

[0004] Each of the documents DE 10 2012 018778 A1, EP 1 936 288 A2 or DE 10 2011 018698 A1 shows a heating system with automatic adaptive hydraulic balancing according to the state of the art.

[0005] The disadvantage of these systems is that these settings, which are calculated based solely on planning data, often do not reflect the reality in the building. For example, the orientation of the room, its use, the size of the radiators, the type of heating valves, and any resulting differences are not taken into account. This often leads to more energy being used than is actually necessary to heat a room to the temperature desired by the user.

[0006] The object of the invention is to provide a heating system in which an adaptive hydraulic balancing is carried out automatically between the consumers, i.e. the radiators. In the heating system according to the invention, no hydraulic balancing with complicated settings on the individual heating valves has to be carried out. Immediately after the installation of the heating valves and the heating system comprising them, the heating valve is adjusted by a computing unit and based on the desired target temperature for the respective room in such a way that an adjusted, i.e. adaptive, setting of the room temperature can be achieved in an optimized manner via the optimized setting of the heating valve position. The balancing is carried out continuously during operation or at defined intervals.

[0007] In the heating system according to the invention with an automatic adaptive hydraulic balancing, the balancing systems are automatically adjusted initially and repeatedly / continuously during operation.

[0008] In connection with the present invention, the following terms are briefly explained and placed in the context of the invention: Comparison

[0009] Setting a continuously acting control device that reacts to changing operating conditions in the hydraulic system (e.g. differential pressure, temperature) during later operation, with the aim of avoiding exceeding the once set target water flow in the design and partial load case Adaptive dynamic balancing

[0010] Automatic determination and continuous or repeated adjustment of a setpoint value (e.g. setpoint water flow) of a continuously acting control device by an algorithm based on measured variables (e.g. differential pressure, flow, temperature), whereby changing operating conditions in the hydraulic system (e.g. differential pressure, temperature) are reacted to during operation, with the aim of avoiding exceeding the setpoint value specified by the algorithm in the design and partial load case. Heating valve

[0011] Hydraulic actuator—e.g., of the room temperature control circuit and / or balancing system—for influencing the flow of media (blocking, allowing, regulating). For the purposes of this invention, actuators with corresponding functionality are referred to as valves, regardless of their technical design.

[0012] The present invention relates primarily to heating systems. The embodiments contained therein—in particular, temperature-based optimization strategies—can be applied in part to cooling and heating-cooling hybrid systems, which are thus equally encompassed by the invention. Such cooling systems include, for example, cooling transfer devices—in particular, surface cooling systems—or hybrid systems that use the same hydraulic system for heating and cooling.

[0013] The heating system according to the invention with automatic adaptive hydraulic balancing of several apartments comprises at least one central heat source, for example a burner, a solar thermal module or comparable devices for generating heat energy. It is also conceivable for heat energy to be provided via a district heating facility that supplies several consumers. The heating system further comprises at least one first consumer, for example a radiator or underfloor heating, each with a local flow branch and a local return branch, and at least one central circulation pump for conveying a heating medium from the heat source via a central flow line to the first consumer. The heating medium can then flow from the first consumer back to the heat source via a central return line.Furthermore, a first heating valve, in particular a thermostatic valve, is provided which is assigned to the first consumer and arranged in the respective flow branch, with a first actuator which is assigned to the first heating valve and is equipped with first control and feedback electronics for setting a valve position value for controlling a volume flow of the heating medium in the first consumer, wherein the actuator preferably acts on an actuating pin of the heating valve.

[0014] The heating system further comprises a detection device for an actual value of the room temperature and a control unit coupled to the detection device and the first actuator. A target value for the room temperature can be stored in the control unit, and a control signal for the valve position value can be output to the control and feedback electronics via the control unit, after which the target value for the room temperature can be adjusted by incrementally actuating the actuator. The control unit can be either a decentralized or centralized unit that is capable of comparing the detected values ​​for the actual temperature with the values ​​for the target room temperature and deriving a control signal from this, which is then output to the control and feedback electronics. The control unit can be provided as a physical element in the room to be heated, in a house, or a building complex.At the same time, it's also possible to provide a virtual control unit, for example, on a server. Cloud solutions are also included.

[0015] In the context of the present invention, the term "target value" for the room temperature refers to the room temperature value that is either set by the user or stored in a temperature profile. This temperature profile can, for example, provide for different room temperatures throughout the day. The heating system sets these temperatures independently and automatically and regularly or continuously checks the actual temperature value in order to compare it with a target value. Once the target value for the room temperature is reached, the respectively set valve position value can be returned to the control unit via the control and feedback electronics. Adaptive hydraulic balancing can thus be carried out by time-controlled, in particular continuous and preferably automatic adjustment of the valve position value.

[0016] The heating system according to the invention thus eliminates the need for initial adjustment of the heating valves. This significantly simplifies the setup and commissioning of the heating system. During operation, the heating system according to the invention is capable of continuously and dynamically responding to temperature changes and can also continuously adjust, monitor, and, if necessary, readjust temperature profiles stored in the control unit. It is also possible to control different temperatures or temperature profiles in different rooms.

[0017] Adaptive balancing systems adjust the flow rate limitation to the current demand. They are characterized by the automatic and continuous or repeated determination of the required temperature target value, depending on the current operating status and demand of all balancing devices / transfer devices, and by pressure-independent control to ensure that the target value is not exceeded. The entire balancing system, including measuring and evaluation devices, is permanently integrated into or connected to the hydraulic system. Temperature-based adaptive hydraulic balancing is primarily based on the measurement and evaluation of media and / or room temperatures.

[0018] A preferred development of the heating system according to the invention provides that at least one second consumer is provided and the control unit outputs control signals for setting the desired value for the room temperature to the control and feedback electronics of the respective consumers for setting the respective valve position value. After the valve position values ​​of the first and second consumers have been compared, the control unit outputs individual valve position values ​​to the control and feedback electronics of the first and second consumers for setting the room temperature. In this way, a comparison can be made between different consumers in the heating system and the adjustment of the individual consumers can be carried out based on this coordination. The system as a whole is therefore considerably more flexible and can also take other influencing factors into account.The control is essentially carried out via the recorded actual temperature in the room and its adjustment with the desired or required target temperature.

[0019] A preferred embodiment of the heating system according to the invention provides that the consumers are arranged in separate rooms and are connected directly to a central control unit or via a control unit assigned to each consumer. In this embodiment, it is possible to use a decentralized room-based control unit that is connected to only one or a few consumers, in addition to a central control unit that combines several consumption units. It is also possible to combine the decentralized control units and provide higher-level control of the heating system of larger complexes via a server-based solution for the control unit(s).

[0020] A special form of the heating system according to the invention is underfloor heating. Here, several valves are arranged on a manifold, which regulate the individual heating circuits of the underfloor heating. The housing is designed so that the valves of the individual heating circuits can calibrate themselves with each other, thus setting the same temperature in all connected heating circuits. For this purpose, additional temperature sensors can be provided on the return or flow lines of the individual heating circuits of the underfloor heating. These sensors enable precise temperature calibration, as further measured values ​​are derived from them, which are passed on to the control unit, where they are used to adjust the valve setting, taking into account the values ​​from the room thermostats or sensors in the building. The control units also enable communication between the individual valves on the manifold.A single control unit can simultaneously control multiple valve actuators in an underfloor heating distributor. A further development of the invention provides for the heating system to be integrated into a distribution network comprising several sections, each of which has independent heating lines, each with at least one heating system. This also provides for synchronization between the heating systems via the control units in the individual heating lines or individual heating systems.

[0021] It is considered advantageous if, in the heating system according to the invention, the distribution network has at least one network control unit for automatic dynamic hydraulic balancing throughout the entire distribution network. The network control unit can control the heating systems individually, i.e., independently of one another, or according to a higher-level balancing, directly or indirectly via the control units of the respective heating systems or via central control units, for example, of individual heating lines.

[0022] Fundamentally, temperature-dependent control of the heating system is achieved according to the present invention. In this context, it proves advantageous if a thermal time response, i.e., a defined room temperature profile, based, for example, on daily or seasonal conditions, can be set in the network control unit, the central control units, and / or the control units of individual units of the heating system using retrievable target value profiles for the room temperature, and if the automatic adaptive hydraulic balancing can be carried out depending on these target value profiles.

[0023] In connection with the previously described heating system design, it proves particularly advantageous if the thermal time behavior is adjusted depending on the time of day, room use, outside temperature, seasonal temperature profiles, room orientation, and / or design. The control unit is capable of comparing the recorded actual temperature with the desired temperature, taking into account the additional influencing factors, and then deriving a valve position value from this and outputting it to the individual heating valves of the respective consumers.

[0024] The heating system according to the invention is characterized in a preferred embodiment in that the actuator is designed as a stepper motor. This stepper motor preferably has a step resolution of between 0 and 1000 steps, in particular between 0 and 800 steps, preferably between 0 and 600 steps, particularly preferably between 0 and 300 steps, and thus allows incremental adjustment with high resolution. The stepper motor can be provided independently of the valve. The stepper motor can also be combined with standard valves and can also be retrofitted to existing valves, thus providing the option of retrofitting existing valves. In an advantageous development of the invention, the heating valve is designed as a stepper motor valve and the actuator is provided as part of the stepper motor valve.

[0025] To control the heating system, it is considered advantageous to provide a target temperature ramp in addition to the target temperature profiles for the room temperature. This ramp specifies the speed within which the target values ​​for the room temperature are reached. The actuation speed of the actuator is also controlled depending on the target temperature ramp. This can be particularly useful when balancing several rooms with different initial or actual temperatures, which should simultaneously reach the required target temperature at a defined time.

[0026] The heating system according to the invention includes a detection device for the room temperature. This can have various configurations. For example, it is conceivable for the detection device to be provided as an independent unit, for example as a measuring sensor or room thermostat. Alternatively or additionally, the detection device can also be arranged directly in the heating valve or in a room thermostat or one or more measuring sensors arranged in the room. The detection device is thus designed either as an integrated or as a stand-alone or independent temperature measuring device and is either part of the control unit or is data-linked to it. In a further embodiment, it is possible for the detection device to also comprise a sensor for room usage. This means that the entire system is only activated when the room is used in a specific, predefined manner.This detects, for example, whether people are present and moving around in the room. Only when a specific activity occurs in the room does the room temperature control system activate. This excludes the possibility that pets moving around the room, for example, might trigger the heating system's activation. Even when people are at rest, for example, during a sleep phase, the heating system is not activated, as the sensor is able to distinguish between the corresponding activities and, based on these and a stored movement profile, sends an activation signal to the heating system.

[0027] Advantageously, the connection of detection devices, actuators, and the respective control unit is provided wirelessly, in particular via Wi-Fi, radio, Bluetooth, or a telephone or mobile network. This can also be done by connecting directly to a router or via a so-called gateway.

[0028] In the heating systems according to the invention, it is provided that the respective network control unit, the control unit and / or the central control unit is arranged decentrally and / or integrated into a data network. For example, it is possible for the respective data acquisition, data conversion and output of control signals to be carried out on a decentralized server, a virtual server or in a web-based cloud Function occurs.

[0029] For remote administration and monitoring, as well as for recording consumption values, it is particularly advantageous if the respective network control unit, the control unit, and / or the central control unit are designed to be operated on a server or virtual server. This allows access to multiple control units from a corresponding control room. It is also possible to network individual control units via this server, so that adaptive hydraulic balancing can be carried out across a larger unit, for example, a building or building complex, or several interconnected heating lines that are supplied, for example, by the same heat source. This also provides improved options for error analysis and troubleshooting. All of this can remotely, that is, via remote maintenance.

[0030] If the control units determine that it is not possible to set the target values ​​for the room temperature with the available heat quantity or flow rate in the heating system, and if this cannot be remedied even through further coordination between the existing heating systems or consumers, it is advantageous if a control signal for the heat source and / or the circulation pump can be output via the respective network control unit, the control unit and / or the central control unit to increase the respective heating or pump output, if required. Conversely, the control signal can of course also be output to reduce, i.e. throttle, the heating or pump output if it is determined that the pump output is too high and therefore less heating medium is required to achieve the respective target values ​​for the room temperature.In another configuration, the heat source, usually the burner, and / or the circulation pump have an interface to the respective control unit. Of course, it is also possible for the heat source and / or the circulation pump to each have their own or shared control units, which are integrated into the heating system and communicate with the overall system. This allows the heat source and / or the circulation pump to output signal values ​​that serve to control the heating system, or the control units can output signals to the heat source or the circulation pump to increase or decrease their output, if necessary within the heating system.

[0031] Equally of inventive significance is a method for performing adaptive hydraulic balancing in a heating system, according to claim 6, in particular for controlling a heating system, in particular a heating system as described above. The method is not limited to heating systems, but can also be used with cooling or air conditioning systems. The invention is not limited solely to heating systems; cooling and air conditioning systems are equally encompassed. The method comprises the steps: a. Recording a current ACTUAL value of the room temperature and passing the ACTUAL value on to a control unit, b. Comparing the ACTUAL value with a TARGET value for the room temperature stored in the control unit, c. Deriving a valve position value from the difference between the TARGET and ACTUAL value, d. Outputting a control signal to control and feedback electronics assigned to a heating valve, e. Actuation by the control and feedback electronics of an actuator of the heating valve to set the valve position value on the heating valve, f. Returning the valve position value to the control unit, g. Repeating steps a.) to f.) until the TARGET value for the room temperature is reached. The target value for the room temperature reflects thermal behavior over time, particularly as a function of the time of day, room usage, outside temperature, seasonal temperature variations, room orientation, and / or design. Based on the recorded actual values ​​and the required target values, the method can be used to predetermine the time required to reach the target value for the room temperature. Based on this value, which is also compared with a corresponding target value, the valve position value can then be output to the control unit. For example, it can be provided that an override, i.e. further opening of the heating valve, takes place first in order to quickly reach the required target value.The actuating speed and the manipulated variable are calculated in the control unit and then output as a control value to the actuator's control and feedback electronics. The control unit also coordinates the current valve position value with the previously calculated or preset value. The control unit uses a corresponding algorithm to continuously adjust and improve the respective values ​​output to the control and feedback electronics. The system is self-learning.

[0032] The method according to the invention is further characterized by the fact that the calibration takes place continuously or at defined time intervals. This means that the respective actual temperature is recorded, for example, at time intervals of 1, 5, and 10 minutes, preferably at time intervals of between 1 and 10 or 15 minutes, or at time intervals adjustable by the user.

[0033] The method according to the invention provides for several interconnected heating valves, and the control unit controls the actuators of the respective heating valves in such a way that the desired value is achieved by individually adjusting the valve position values ​​of the respective heating valves. In particular, the respective heating valves are opened and / or closed independently, but coordinated across the entire heating system. The independent opening and / or closing preferably occurs stepwise.

[0034] An advantageous development of the method provides that the controller or the respective control unit takes into account different loads occurring at the heating valves, in particular pressures, during the balancing process. The dimensions of the respective radiator are also taken into account. The method is able to detect under- or over-dimensioning of the radiator and report this back to the control unit. Detection is carried out via the response behavior of the respective radiator valve and the time until a target temperature value is reached. The respective values ​​for the radiators or heat consumers can be stored in the control unit and taken into account as an influencing factor when adjusting the balancing behavior. In particular, it can be provided that the pressure in the heating system is automatically regulated indirectly via the temperature control.

[0035] In a further embodiment of the method according to the invention, temperature ramps for reaching the target temperature are stored in the control unit, and the actuators are controlled depending on the temperature ramps. Of course, other factors influencing the room temperature are also taken into account.

[0036] In the method according to the invention, the actuator is preferably actuated step by step with a step resolution of between 0 and 1000 steps, in particular between 0 and 800 steps, preferably between 0 and 600 steps, particularly preferably between 0 and 300 steps.

[0037] An advantageous development of the method provides for adaptive hydraulic balancing in a building heating system equipped with a distribution network. The aforementioned method steps a.) to f.) are performed in individual sections of a distribution network, and control signals are calculated and output for these sections. Balancing for the entire network is preferably performed via a network control unit.

[0038] In this context, it is considered advantageous if the distribution network has multiple control units and / or central control units, and a data connection between the network control unit and the multiple control units and / or central control units is provided. This makes it possible for multiple control units or central control units that control individual heating lines to be combined into one system and controlled across all units by a single network control unit. This allows for balancing not only within the same heating line, the same residential unit, or the same building unit, but also across a larger building complex.In this context, the process not only simplifies control and implements adaptive hydraulic balancing, but also allows for a corresponding shift in energy demand, which leads to a reduction in overall energy demand. Since heat energy only needs to be provided when needed, the burner and pump outputs can be controlled selectively and based on demand. This prevents idle times or excess capacity.

[0039] In this context, it proves to be particularly advantageous if the distribution network and / or the sections have a plurality of heating lines, in particular supplied by a central heat source, and the at least one network control unit and / or a plurality of control units and / or central control units carry out or carry out the balancing across the heating lines.

[0040] The network control unit outputs the control signals, either directly or via the control units and / or central control units, to the control and feedback electronics of the heating valves. Depending on the load required to supply all consumers evenly with heat energy, the central control units can perform a balancing operation by comparing several heating valves and the pressures or loads present there. This makes it possible to optimize the heating control system as a whole and, if necessary, adjust it so that no excess energy is consumed. Overall, the heating system can be operated more economically, efficiently, and in a more environmentally friendly manner thanks to the method according to the invention.

[0041] It is advantageous if the respective network control unit, the control units and / or the central control units issue a control signal for the heat source and / or the circulation pump to increase or decrease the heating and / or pumping output as required in order to achieve the target value and depending on the respective load. This includes the option of either the network control unit or the central or decentralized control units communicating directly with the pump or boiler control. During the balancing process, the demand from various sub-units can be bundled in the control units and passed on to the respective providers of heat energy or the associated pumps. An initial attempt can therefore be made to cover the different loads in the system by controlling the heating valves accordingly.If this is not successful enough to set the desired setpoint values ​​for the room temperature, the control units then send the relevant signals to the boiler or pump control, thereby increasing their output so that all requirements from the control units are taken into account and a setpoint value for the room temperature can be set as intended. The control signal to the boiler or pump control takes into account the consumers present in the system, as well as the respective temperature profiles or requested setpoint temperatures. It is also possible to control different temperatures or temperature profiles in different rooms and to distribute the available amount of heating medium in such a way that the setpoint temperatures in the rooms are quickly reached and maintained for the respective period of time.Pump(s) can have their own control units that communicate with the overall system. However, it is also possible for the pump(s) or boiler to have an interface, allowing them to be integrated into the heating system and connected to the control units provided there.

[0042] At the same time, however, additional parameters such as the time of year, the orientation of the respective room, the time of day or the size of the rooms to be heated or the dimensions of the available radiators are also taken into account. The algorithm stored in the control units uses these values ​​to calculate the control signals for adjusting the heating valves and thus adapting the opening and closing widths of the valves and thus the flow rate and flow rate of the heating medium. By comparing all parameters, the heating system can be optimized. In addition to avoiding peak loads, this also saves energy, as the heating system can be easily adjusted as needed.

[0043] In this context, it proves advantageous if the control signals are forwarded wirelessly, in particular via a Wi-Fi connection, radio, Bluetooth, or a telephone or mobile network. The respective network control unit, the control units, and / or the central control units are provided in a decentralized manner and / or are integrated into a data network. Alternatively, all control units can also be operated on a server or virtual server, enabling web-based or cloud-based implementation of the method and thus control of a heating system. In this context, it proves advantageous if the respective network control unit, the control unit, and / or the central control unit are operated on a server or virtual server.

[0044] All relevant parameters of the heating system are continuously recorded and stored. In this context, it is planned that information about the function of the heating system, such as usage data, performance data and consumption values, is shown on a display device and output to a user so that the user can monitor the heating system. The function is also provided that the parameters of the heating system can be adjusted by a user via a software product, e.g. an app or in a browser interface and fed back into the heating system. In addition, faults, malfunctions or failures can be reported to the user and displayed via a graphical interface. This also includes warning messages such as leaks in the heating system, which are triggered, for example, by a burst pipe or other damage to the pipes.The control units can also be used to integrate the heating system into a so-called smart home environment, i.e., a digital building monitoring or control system. This allows the heating system to be controlled as a component of a higher-level building control system. In this context, synchronization between browsers, desktop devices, and mobile devices, such as cell phones or tablet computers, is also conceivable and planned. This provides the user with the ability to monitor and adjust the heating system at any time and from any location.

[0045] The method according to the invention also includes the simultaneous control of several heating systems and / or distribution networks via a server or a network.

[0046] The heating system according to the invention is preferably designed as a self-learning system. Therefore, the invention also includes a method for training an artificial neural network. (KNN), in particular for controlling a heating system and for modifying the method described above, which predicts at least one valve position value for achieving a target value of a room temperature in a heating system. The behavior of the KNNis characterized by a set of parameters, which can, for example, contain the weights with which the inputs supplied to a neuron and / or another computing unit are calculated to activate the neuron. In the multi-step method, a set of learning data sets is first provided, with each learning data set for a valve position value comprising a set of parameters that characterizes the achievement of a target value of a room temperature and at least one measure for the incremental actuation of the actuator, which leads to the achievement of an opening value of the heating valve. KNNFor each learning data set, the set of parameters contained therein and the measure for the incremental actuation of the actuator contained therein are fed as inputs in order to obtain a prediction and / or classification for at least one opening value of the heating valve as output. The prediction and / or classification for the opening value of the heating valve is then compared with the value for the opening value of the heating valve contained in the learning data set. Furthermore, an incremental actuation of the actuator is evaluated, which depends on a deviation determined in the comparison. The parameters and / or the learning data sets of the ANN are adapted with the optimization goal of improving the value of the incremental actuation of the actuator. This predicts or makes predictable the response to changed environmental parameters.The appropriately trained ANN can thus anticipate changes resulting from variations in the respective input parameters and thus react in advance to the changed conditions in order to achieve the target values ​​more quickly and accurately.

[0047] In the method according to the invention, the learning data set additionally comprises an identification of the installation location of the heating valve(s) in a heating system and / or an identification of the position of the heating system in a heating circuit or a higher-level heating system and / or a measure of the ambient temperature at the time the opening value of the heating valve is determined and / or information as to how the heating system is constructed and / or information as to how the load in the heating system behaves and / or at least excerpts of planning data for the heating circuits or the building or room to be heated and / or information about the heating valves present in the heating system and coupled to one another and / or information as toThe proportion of a total heat quantity or pressure increase in the heating system that is or can be made available at the respective heating valve, and / or information regarding the maximum output of the burner or heat pump assigned to the heating system. These parameters of the learning data set can be fed into the ANN as additional inputs, thus optimizing the training performance during heating system operation.

[0048] The invention also comprises an artificial neural network (KNN) trained by the method as described above and a data set characterizing the KNN obtained as a result of the aforementioned method.

[0049] Furthermore, the invention comprises a method based on the ANN for predicting and / or classifying the valve control value of a respective heating valve in a heating system. A set of parameters characterizing the required target room temperature is determined; at least one measure of the valve control value of each heating valve in a heating system is determined; and these parameters and the measure of the valve control value of a respective heating valve in a heating system are fed to the trained ANN as inputs. Finally, at least one prediction and / or classification for a measure of the valve control value for achieving the target temperature in the respective heating system or heating circuit is retrieved as output from the ANN and output via the control unit to the heating valves or the control and feedback electronics provided there.This then activates the corresponding actuator to adjust the calculated valve control value. Starting from random starting values, for example, the ANN's parameters evolve over the course of such training, so that the ANN more or less accurately predicts the valve control value of a particular heating valve for the parameters contained in each learning data set in conjunction with the respective target temperature value, and then implements this value through the control units.

[0050] In the aforementioned method, the ANN is preferably additionally supplied with an identification of the location at which at least one heating valve was installed and / or a measure of the ambient temperature at the location where the heating valve was installed as inputs.

[0051] In the method according to the invention, in response to the fact that the target temperature value is not or cannot be reached with a valve control value of a first heating valve predicted for a first installation location, a further prediction and / or classification for a valve control value of a second heating valve provided at a second installation location is retrieved as an output from a trained ANN and output as an adjustment value. Thus, coordination or adjustment between the various heating valves can be carried out depending on the situation and independently of stored temperature profiles. These temperature profiles can also be adjusted and thus optimized based on the calculated values.

[0052] Overall, this means that the experience gained in the learning data sets regarding the valve control value of a particular heating valve can be used to make accurate predictions for the future setting of the valve control value. If the amount of learning data sets is sufficiently large and diverse, the ANN can generalize the findings contained therein to the extent that it can subsequently predict the specific setting of the valve control values ​​in a heating system with sufficient accuracy, even for completely unknown combinations of target values ​​and ambient conditions. The generalization ability of ANNs is particularly important in the context of heating system operation. If a similar building with the same configuration and arrangement of heat sources is constructed at a different location, it is very likely that the automatic, adaptive hydraulic balancing can be carried out more quickly and accurately.

[0053] The invention also relates to one or more computer programs containing machine-readable instructions which, when executed on one or more computers and / or on one or more control units and / or one or more servers, cause the one or more computers and / or the one or more control units and / or the one or more servers to execute a method as described above. The computer program(s) can be embodied, for example, on a machine-readable data carrier or a downloadable product that can be acquired and loaded over a network.

[0054] The subject matter of the invention is explained below with reference to figures, without the subject matter of the invention being limited thereby.

[0055] It shows: Fig. 1: Embodiment of a heating system arranged in a building, Fig. 2:Example of a heating valve, Fig. 3: an example of the coordination of several heating valves to achieve a target temperature and Fig. 4: a special application of a heating system.

[0056] Fig. 1 shows a building 100, which has a total of three separate heating lines 20a, b, c. In the building 100, a heat generator 11 is provided, which provides the heat required for the individual consumers V1. Pumps 21a, b, c are respectively assigned to the heating lines 20a, b, c. Each of the heating lines 20a, b, c supplies a number of apartments 30a-l with heat. In the apartments 30a1, the consumers V1 are arranged, shown as radiators in only one of the apartments 30b by way of example. Each of these consumers V1 is assigned a control unit 40a-l. This is coupled to a heating valve HV1. In the heating valve HV1, a detection device EV (cf. Fig. 2) for the actual temperature in the respective room.

[0057] The detection device EV records the current actual value of the room temperature and transmits it to the control unit 40a-l. Target values ​​for the room temperature are stored in the control unit 40a-l. The control unit 40a-l compares the measured actual value with the stored target value and derives a valve position value from this. This valve position value is then output to the heating valve HV1, which then uses the control and feedback electronics RE provided in the heating valve HV1 (see Fig. 2 ) an adjustment of an actuator S1 which is also provided in the heating valve HV1 (cf. Fig. 2). This either throttles or increases the volume flow of the heating medium supplied via the flow line 3, ultimately setting the desired setpoint temperature. The control unit 40a-l calculates the valve position value as a derivative of the deviation between the setpoint and actual values.

[0058] The actuator S1 of the heating valve HV1 is operated via the control and feedback electronics RE to adjust the valve position. Once the calculated valve position is reached, it is returned to the control unit 40a-l and stored there. The previously described steps are repeated until the desired room temperature is reached. Continuous adjustment within adjustable time intervals thus ensures that the desired temperature is maintained in the respective room.

[0059] As from Fig. 1As can be seen, several control units 40a-l are arranged in each heating circuit 20a, b, c, each of which acts on the corresponding consumers V1 or is coupled to them. In the heating system 10, the various control units 40a-l are coordinated or adjusted, and thus all consumers V1 in the heating circuits 20a, b, c are adjusted. The consumers V1 of the individual heating circuits 20a, b, c can thus be controlled depending on the temperature, so that the amount of heat provided can be optimally distributed to the consumers V1. A uniform temperature profile is thus achieved in all rooms. The control units 40a-l can, if necessary, also access control units 41a, b of the pumps 21a, b, c or the heat generator 11 and adjust or throttle the pump output or the amount of heat provided as needed. All consumers V1 of a heating circuit 20a, b, c are connected to the flow 3 orReturn line 4 is connected to each other. The control units 40a-l can also be connected to a central server 50, which handles the system-wide balancing and the corresponding control of the heating valves HV1. The network control unit NE thus provided processes the values ​​recorded by the respective control units 40a-l of the heating lines 20a, b, c and calculates the valve position values ​​to be output, which are then implemented via the control units 40a-l in the respective building units or apartments 30a-l. By bundling the information from the control units 40a-l of the individual building units, an overall balancing can be carried out across the building in a simple manner via the network control unit NE, and the output or load at the individual consumers V1 can be optimized and balanced in such a way that an automatic, adaptive hydraulic balancing can be carried out in the entire heating system 10.This can be done dynamically and without mechanical balancing on individual heating valves HV1. The output of these elements can be included in the balancing or the output can be adjusted via the control units of the pumps 21a, b, c or the boiler 11. Alternatively, an interface 42 can be provided on the pumps 21a, b, c and / or the boiler 11 for integration into the heating system 10 via the control units 40a-l or the network control unit NE.

[0060] Other influencing factors when carrying out the adaptive hydraulic balancing can include, for example, the installation location of the respective heating valve HV1 or the consumer V1 in the heating system 10, the position of the heating system 10 in a heating circuit 20a, b, c or a higher-level heating system and / or a measure of the ambient temperature at the time the opening value of the heating valve HV1 is determined. In addition, information regarding how the heating system 10 is constructed can also be taken into account. It is also possible to take into account information regarding how the load occurs in the heating system 10 and / or at least excerpts of planning data for the heating circuits 20a, b, c or the building 100 to be heated or the room to be heated and / or information about the heating valves HV1 present and coupled to one another in the heating system 10.The calculation of the valve position value necessary to achieve a target temperature TS can also take into account the proportion of a total heat quantity or a pressure build-up in the heating system 10 that is available or can be made available at the respective heating valve HV1 and / or information as to the maximum output that the burner 11 assigned to the heating system 10 or the heat pumps 21a, b, c assigned to the heating system can provide.

[0061] Fig. 2shows a design of a heating valve HV1. In the exemplary embodiment, this has, in addition to the actuator S1 for setting a valve position value for controlling a volume flow of a heating medium to or from the first consumer V1, a detection device EV for the actual value of the room temperature. In addition, control and feedback electronics RE are arranged in the heating valve HV1, which is connected to the control unit 40a-k and receives control signals from it for adjusting the actuator S1. This actuator S1 acts on an actuating pin of the heating valve HV1, via which the opening and closing width of the heating valve HV1 is set and thus the volume flow of the heating medium is controlled. The control and feedback electronics RE is able to return the currently set valve position value to the control unit 40a-k, where it is then used for further calibration.In the embodiment shown, the actuator S1 is . Fig. 2designed as a stepper motor with a step resolution of between 0 and 1000 steps, in particular between 0 and 600 or 0 and 300 steps. The amount of heating medium made available to consumer V1 can be adjusted by actuating the actuator S1. Due to the continuous feedback between the heating valve HV1 and the control unit 40a-k, a permanent hydraulic balance can be carried out in the respective consumer V1 and, through higher-level recording and processing of the respective valve position values, an adaptive hydraulic balance can be carried out in an entire heating system, so that the actuators S1 are always optimally controlled and actuated. This ultimately leads to the target temperature TS or temperatures being reached in the entire heating system 10. It is also possible to use different temperatures orTo control temperature profiles in different rooms and to distribute the available amount of heating medium in such a way that the target temperatures TS in the rooms are quickly reached and maintained over the respective planned period.

[0062] Fig. 3shows an example of the coordination of several heating valves HV1 to achieve a target temperature TS. The required target temperature TS is 22°C in the exemplary embodiment. The detection devices EV provided in the individual rooms have determined different actual temperatures TI 1, TI 2, TI 3, TI 4 in the rooms at the first measurement time MZ. The control unit 40a-k now adjusts the valve position values ​​of the heating valves HV1 in the individual rooms such that the target temperature TS is reached simultaneously in all rooms by a target time TZ. For example, in the room with the actual temperature TI 4 at the first measuring time MZ, which in the exemplary embodiment has the lowest temperature, a larger valve opening is set than in the room with the actual temperature TI 1 at the first measuring time MZ, which is already close to the target temperature TS.Through continuous feedback and comparison between the individual rooms and the setting values ​​for the heating valve HV1 assigned to the consumers V1 there, automatic control can be achieved such that the amount of heat provided is distributed among the rooms in such a way that they have the same target temperature TS at a target time TZ. This also takes into account the orientation of the rooms, their use and size, as well as the dimensioning of the radiators in the respective rooms. These parameters are ultimately decisive for achieving the target temperature TS and the time required to achieve it. The valve position values ​​of the heating valve HV in the individual rooms are stored in the control unit 40a-k and continuously compared with the recorded room temperatures.If necessary, the valve position values ​​can be adjusted, thus adjusting and controlling the volume flow in the heating valve HV1 to the respective consumers V1. The heating system 10 according to the invention also allows the temperature once reached in the rooms to be maintained over a maintenance period HZ by continuously recording the temperature and, if necessary, readjusting the valve position values. The recording device EV records the actual temperature of the room continuously or at defined time intervals ZI 1 - ZI 4 of, for example, 1, 5, and 10 minutes, preferably between 1 and 10 or 15 minutes, or at time intervals adjustable by the user.

[0063] A special application of a heating system 10 in a building with several apartments 30a-l, which is not part of the invention, is Fig. 4While in conventional systems with several heating lines 20a, b, c each heating line 20a, b, c is assigned its own circulation pump 21, Fig. 4 the special case in which three heating lines 20a, b, c of different lengths are provided, which have a common pump 21. In this example, digital line regulating valves 50a, b, c are provided in the heating lines 20a, b, c, which also communicate with control units 40a, b, c and receive control signals from them.

[0064] The heating system 10 or the control unit 40a, b, c assigned to it first simulates a heating system 10 with equally long lines (shown as dashed lines) and automatically performs an adjustment via the control of the digital line regulating valves 50a, b, c. This achieves a uniform pressure distribution and a corresponding load on the consumers (not shown) in the individual apartments 30a-l throughout the system. In the next step, the room temperature is then set using the method or heating system 10 as described above. This involves coordination between the control units 40a, b, c, the respective consumers (not shown) and the line regulating valves 50a, b, c and the pump 21, and to create uniform conditions throughout the heating system 10. The control or adjustment between the individual units or apartments 30a-l then takes place in the manner described above.

Claims

1. Heating system (10) with an automatic adaptive hydraulic adjustment: • at least one central heating source, for example burner (11) • at least one first consumer (V1), each with a local supply branch (3) and a local return branch (4), • at least one central circulation or heat pump (21a, b, c) for conveying a heating medium from the heating source (11) via a central supply line (3) to the first consumer, wherein the heating medium can flow back from the first consumer (V1) to the heating source via a central return line (4), • a first heating valve (HV1), in particular a thermostatic valve, assigned to the first consumer (V1) and arranged in the respective flow branch (3), • a first actuator (S1) associated with the first heating valve (HV1) and equipped with first control and feedback electronics (RE) for setting a valve position value for controlling a volume flow of the heating medium in the first consumer (V1), wherein the actuator (S1) preferably acts on an actuating pin of the heating valve (HV1), • a detection device (EV) for an ACTUAL value of the room temperature (TI1-4), • a control unit (40a-l) coupled to the detection device (EV) and the first actuator (S1), wherein a setpoint value for the room temperature (TS) can be stored in the control unit (40a-l) and a control signal for the valve position value can be output to the control and feedback electronics (RE) via the control unit (40a-l) and the setpoint value for the room temperature (TS) can be set by incremental actuation of the actuator (S1) and, when the setpoint value for the room temperature (TS) is reached, the respective set valve position value can be transmitted to the control unit (40a-l) via the control and feedback electronics (RE).value for the room temperature (TS) can be set and the respective set valve position value can be fed back to the control unit (40a-l) via the control and feedback electronics (RE) when the set value for the room temperature (TS) is reached and the adaptive hydraulic balancing can be performed by time-controlled, in particular continuous, adjustment of the valve position value by means of the control unit (40a-l), in particular continuous adjustment of the valve position value, characterised in that integration of the heating system into a distribution network comprising a plurality of sections is provided, the sections having independent heating lines (20a, b, c) each with at least one consumer (V1), the heating lines each supply a number of flats (30a-l) and the distribution network has at least one network control unit (NE) for dynamic hydraulic balancing in the entire distribution network, wherein the network control unit (NE) controls the heating system directly via the control units (40a-l) or via the central control units.

2. Heating system (10) according to claim 1, characterised in that in that at least one second consumer is provided and the control unit (40a-l) for setting the setpoint value for the room temperature (TS) outputs control signals for setting the respective valve position value to the control and feedback electronics (RE) of the respective consumers (V1) and, in the control unit (40a-l), after the valve position values of the first and second consumers (V1) have been equalised, an output of individual valve position values to the control and feedback electronics (RE) of the first and second consumers (V1) is provided for setting the room temperature and the consumers (V1), in particular in the control and feedback electronics (RE) of the first and second consumers (V1), are provided for setting the room temperature, and feedback electronics (RE) of the first and second consumer (V1) is provided for setting the room temperature and the consumers (V1) are arranged in particular in separate rooms and a coupling is provided directly with a central control unit or via a control unit (40a-l) respectively assigned to the consumers and wherein in the network control unit (NE), the central control units or the control units (40a-l), a thermal time behaviour can preferably be stored via retrievable setpoint value profiles for the room temperature and the adaptive hydraulic balancing can be carried out as a function of the setpoint value profiles for the room temperature and an adaptation of the thermal time behaviour is provided dependent of the time of day, the room use, the outside temperature, the seasonal temperature curves, room orientation and / or design.

3. Heating system (10) according to one of the preceding claims, characterised in that the actuator (S1) is designed as a stepper motor which has a step resolution of between 0 and 1000 steps, in particular of between 0 and 800 steps, preferably of between 0 and 600 steps, particularly preferably of between 0 and 300 steps, wherein the heating valve (HV1) is designed in particular as a stepper motor valve and the actuator (S1) is provided as part of the stepper motor valve.

4. Heating system (10) according to one of the preceding claims, characterised in that, in addition to the setpoint value profiles for the room temperature, a setpoint temperature ramp is provided and an actuating speed of the actuator (S1) is provided as a function of the setpoint temperature ramp and / or the detection device (EV) is preferably designed as an independent temperature measuring device or as a temperature measuring device integrated in the heating valve (HV1) or in a room thermostat.

5. Heating system (10) according to one of the preceding claims, characterised in that the respective network control unit (NE), the control unit (40a-l) and / or the central control unit are arranged decentrally and are integrated into a data network or the respective network control unit (NE), the control unit (40a-l) and / or the central control unit are designed to be operable on a server (50) or virtual server and are transmitted via the respective network control unit (NE), the control unit (40a-l) and / or the central control unit, the control unit (40a-l) and / or the central control unit is designed to be operable on a server (50) or virtual server and a control signal for the heating source and / or the circulation or heat pump (21a, b, c) can be output as required via the respective network control unit (NE), the control unit (40a-l) or the central control unit in order to increase or reduce a respective heating or pump output.

6. Method for carrying out an automatic, adaptive hydraulic balancing in a heating system (10) according to one of claims 1 to 5, comprising the steps of: a. Detection of a current ACTUAL value of the room temperature (TI 1-4) and transmission of the ACTUAL value of the room temperature (TI 1-4) to a control unit (40a-l), b. Comparison of the ACTUAL value of the room temperature (TI 1-4) with a SET value for the room temperature (TS) stored in the control unit (40a-I), c. Derivation of a valve position value from the deviation between the SET value for the room temperature (TS) and the ACTUAL value for the room temperature (TI 1-4), d. Output of a control signal to a control and feedback electronics (RE) assigned to a heating valve (HV1), e. Actuation of an actuator (S1) of the heating valve (HV1) by the control and feedback electronics (RE) to set the valve position value on the heating valve (HV1), f. Return of the valve position value to the control unit (40a-l), g. Repeating steps a.) to f.) until the target value for the room temperature (TS) is reached, wherein in particular the setpoint value for the room temperature (TS) represents a thermal time behaviour, in particular as a function of the time of day, the room use, the outside temperature, the seasonal temperature curves, the room orientation and / or design, and / or the adjustment takes place continuously or at defined time intervals (ZI1-4) and / or preferably a plurality of interconnected heating valves (HV1) are provided, and the control unit (HV2) controls the actuators (S1) of the respective heating valves (HV1) in such a way that the setpoint value for the room temperature (TS) is achieved by individual adjustment of the valve position values of the respective heating valves (HV1), wherein an independent opening and / or closing of the respective heating valves (HV1) is carried out and / or the control system takes into account different load incidents occurring at the heating valves (HV1) during balancing, characterised in that an adaptive hydraulic balancing is carried out in a building heating system equipped with a distribution network comprising several sections, wherein the sections have independent heating lines (20a, b, c) each with at least one consumer (V1) and the heating lines each supply a number of flats (30a-l) and wherein the steps a.) to f.) are each carried out in individual sections of a distribution network and the balancing for the entire network is carried out via a network control unit (NE), wherein the distribution network in particular has a plurality of control units (40a-l) or higher-level central control units and a data coupling of the network control unit (NE) with the plurality of control units (40a-l) and / or central control units is provided and wherein the distribution network or the sections have a plurality of heating lines (20a, b, c) supplied in particular by a central heating source, and the at least one network control unit (NE) and a plurality of control units (40a-l) or central control units perform the balancing across the heating lines (20a, b, c) and the network control unit (NE) outputs the control signals either directly or via the control units (40a-l) and / or central control units to the control and feedback electronics (RE) of the heating valves (HV1), wherein the network control unit (NE), the control units (40a-l) or the central control units in order to achieve the setpoint values for the room temperature (TS) and, depending on the load required in each case, outputs a control signal for the heating source and / or the circulation or heat pump(s) (21a, b, c) as required in order to increase or reduce a heating and / or pump output.

7. Method according to claim 6, characterised in that temperature ramps for achieving the setpoint values for the room temperature (TS) are stored in the control unit (40a-I) and the actuators (S1) are controlled as a function of the temperature ramps and / or the actuator (S1) is actuated in steps with a step resolution of between 0 and 1000 steps, in particular between 0 and 800 steps, preferably between 0 and 600 steps, particularly preferably between 0 and 300 steps8. Method according to one of claims 6 or 7, characterised in that the control signals are forwarded wirelessly, in particular via a W-LAN connection, radio, Bluetooth or a telephone or mobile network and / or the network control unit (NE), the control units (40a-l) and / or the central control units are provided decentrally or are integrated into a data network or the network control unit (NE), the control unit (40a-l) and / or the central control unit are operated on a server (50) or virtual server, whereby the simultaneous control of several heating systems (11) and / or distribution networks takes place via a server (50) of the network.

9. A computer-implemented method for training an artificial neural network (ANN) for executing an adaptive hydraulic adjustment according to any one of claims 6 to 8 in a heating system according to any one of claims 1 to 5, which predicts at least one valve position value for achieving a setpoint value of a room temperature (TS) in a heating system (10), wherein an integration of the heating system into a distribution network comprising a plurality of sections is provided and the sections comprise independent heating lines (20a, b, c) each having at least one consumer (V1) and the heating lines each supplying a number of flats (30a-l), the behaviour of the ANN being characterised by a set of parameters, with the steps of • a set of learning data sets is provided, wherein each learning data set for a valve position value comprises a set of parameters characterising the achievement of a setpoint value of a room temperature (TS), at least one measure for the incremental actuation of the actuator (S1), which is required to achieve an opening value of the heating valve (HV1); • For each learning data set, the set of parameters contained therein and the measure for the incremental actuation of the actuator (S1) contained therein are fed to the ANN as inputs in order to obtain a prediction and / or classification for the at least one opening value of the heating valve (HV1) as an output; • the prediction and / or classification for the opening value of the heating valve (HV1) is compared with the value for the opening value of the heating valve (HV1) contained in the learning data set; • An incremental actuation of the actuator (S1) is evaluated, which depends on a deviation determined in the comparison; • the parameters and / or the learning data sets of the ANN are adapted with the optimisation objective of improving the value of the incremental actuation of the actuator (S1) and whereby the learning data set optionally additionally contains • an identification of the installation location of the heating valve (HV1) in the heating system (10) and / or • an identification of the position of the consumer (V1) in a heating line (20a, b, c) and / or • a measure of the ambient temperature at the time the opening value of the heating valve (HV1) is set, and / or • an information on how the heating system (10) is constructed, and / or • an information on the load behaviour in the heating system (10) and / or • at least excerpts of planning data for the heating lines (20a, b, c) or the building (100) to be heated, a number of flats (30a-l) or the room to be heated, and / or • an information about the heating valves (HV1) present in the heating system (10) and coupled to one another, and / or • an information on the proportion of a total heat quantity or a pressure in the heating system (10) that is available at the respective heating valve (HV1) and / or • an information on the maximum output that can be provided by the heat generator assigned to the heating system (10), in particular the burner (11) or the circulation or heat pump (21a, b, c) assigned to the heating system (10), as further inputs to be supplied to the CN.

10. Artificial neural network, ANN, trained with the method according to claim 9.

11. Data set of parameters characterising a ANN obtained by the method according to claim 9.

12. Computer-implemented method for predicting and / or classifying the valve setting value of a respective heating valve (HV1) in a heating system (10) according to any one of claims 1 to 5, wherein the integration of the heating system into a distribution network having a plurality of sections is provided with the sections having independent heating lines (20a, b, c) each having at least one consumer (V1), with the heating lines (20a, b, c) each supplying a number of flats (30a-l), with the steps: • A set of parameters characterising the required SET value of the room temperature (TS) is determined; • at least one measurement for the valve control value of each heating valve (HV1) in the heating system (10) is determined; • the parameters and the measure for the valve actuating value of a respective heating valve (HV1) in the heating system (10) are fed to a trained ANN as inputs; at least one prediction and / or classification for a measure of the valve setting value for reaching the setpoint value of the room temperature (TS) in the heating system (10) or heating line (20a, b, c) is retrieved as an output from the ANN, wherein optionally the ANN is additionally provided with the following information: • an identification of the location where at least one heating valve (HV1) was installed, and / or • a measure for the ambient temperature or ACTUAL value of the room temperature (TI1-4) at the location of installation of the heating valve (HV1) is supplied as input and / or preferably in response to the fact that the SET value of the room temperature (TS) is not reached with a valve setting value of a first heating valve (HV1) predicted for a first installation location, a further prediction and / or classification for a valve position value of a second heating valve (HV1), which is provided at a second installation location, is retrieved as an output from a trained ANN.

13. One or more computer programs comprising machine-readable instructions which, when executed on one or more computers and / or on one or more control units (40a-l) and / or on one or more servers (50), cause the one or more computers and / or the one or more control units (40a-l) and / or the one or more servers (50) to execute a method according to claim 12.

14. Machine-readable data carrier and / or download product comprising the one or more computer programs according to claim 13.